Download CISSP Summary V1.1

Maarten de Frankrijker, CISSP
Revised by Christian Reina, CISSP
CISSP summary
Version 1.1
This document may be used only for informational, training and noncommercial purposes. You are free to copy, distribute, publish and alter this document under the conditions that you give credit to the original author.
2009 - Maarten de Frankrijker, CISSP. Revised by Christian Reina, CISSP.
Concepts
Security policies, standards and guidelines
Legislative drivers
CIA Negative: (DAD disclosure alteration and destruction)
Confidentiality prevent unauthorized disclosure
Integrity no unauthorized modifications, consistent data
Availability reliable and timely accessible
Policies first and highest level of documentation
FISMA(federal agencies)
Phase 1 categorizing, selecting minimum controls, assessment
Phase 2: create national network of secures services to assess
NIST
8 elements reassessments owners have responsibilities. Benefits:
consistent; comparable; repeatable
OECD
accountability, awareness, ethics, etc loads of one word things
Identification user claims identity, used for user access control
Authentication testing of evidence of users identity
Accountability determine actions to an individual person
Authorization rights and permissions granted
privacy level of confidentiality and privacy protections
Controls
Prime objective is to reduce the effects of security threats and
vulnerabilities to a tolerable level
Risk analysis process that analyses threat scenarios and
produces a representation of the estimated Potential loss
Types Physical, Technical and Administrative
Domain 1 - Security Management
Information classification
WHY? Not all data has same value, demonstrates business
commitment to security, Identify which information is most
sensitive and vital
Criteria Value, age, useful life, personal association
Levels
Government, military
Unclassified
Sensitive but unclassified (answers to test, Healthcare)
Confidential (some damage)
Secret (Serious damage)
Top Secret (Grave damage)
Private sector
Public
Sensitive
Private
Confidential
Security Awareness
Technical training to react to situations, best practices for Security
and network personnel
Employees, need to understand policies then use presentations
and posters etc to get them aware
Losses
staff members pose more threat than external hackers
loss of money stolen equipment,
loss of time work hours
loss off reputation declining trusts and
loss of resources bandwidth theft
Very first is called Senior management Statement of Policy,
Stating importance, support and commitment
Types
Regulatory (required due to laws, regulations,
compliance and specific industry standards!)
Advisory (not mandatory but strongly suggested)
Informative to inform the reader
Information policy has classifications and defines level of access
and method to store and transmit information
Security policies has Authentications and defines technology
used to control information access and distribution
SYSTEM security policy lists hard software to be used and steps
to undertake to protect infrastructure
Standards Specify use of specific technologies in a uniform way
Guidelines same as standards but not forced to follow
Procedures Detailed steps to perform a task
Baseline Minimum level of security
Security planning
Security Planning involves security scope, providing security
management responsibilities and testing security measures for
effectiveness. Strategic 5 years Tactical shorter than strategic
Operational day to day, short term
Roles and responsibilities
Senior Manager ultimate responsibility
Information security Officer functional responsibility
Security Analyst Strategic, develops policies and guidelines
Owner
Responsible for asset
Determine level of classification
Review and change classification
Can delegate responsibility to data custodian
Authorize user privileges
Custodian
Run regular backups/restores and validity of them
Insuring data integrity and security (CIA)
Maintaining records in accordance to classification
Applies user authorization
End-user
Uses information as their job
Follow instructions in policies and guidelines
Due care (prevent open view by e.g. Clean desk)
Use corporation resources for corporation use
Auditor examines security controls
Risk Management
GOAL
Determine impact of the threat and risk of threat occurring
ACTIVITIES
Primary (risk assessment, mitigation methodology)
Secondary (data collection and sources for risk analysis)
Types of Risk
Inherent chance of making an error with no controls in place
Control chance that controls in place with prevent, detect or
control errors
Detection chance that auditors won’t find an error
Residual risk remaining after control in place
Business concerns about effects of unforeseen circumstances
Overall combination of all risks aka Audit risk
Preliminary Security Examination (PSE): Helps to gather the
elements that you will need when the actual Risk Analysis takes
place.
ANALYSIS
Steps: Identify assets, identify threats, and calculate risk.
Qualitative HAPPY FACES
Higher level , brainstorming, focus groups etc
Quantitative VALUES!!
SLE (single Loss Expectancy) = Asset Value * Exposure
factor (% lost of asset)
ALE (Annual loss expectancy) = SLE * ARO
(Annualized Rate of occurrence)
Remedies: Accept, mitigate(reduce by implementing controls
calculate costs-), Assign (insure the risk to transfer it), Avoid (stop
business activity)
Loss= probability * cost
Risk Based Audit approach
Planning and information gathering
Access internal controls
Compliancy testing
Substantive tests
Finalize the audit
Access control
ACCESS is flow of information between a subject and an object
CONTROL security features that control how users and systems
communicate and interact with other systems and resources
Subject is active entity that requests access to an object or data
within the object (user, program)
Object is a passive entity that contains information (computer,
database, file, program)
access control techniques support the access control models
Domain 2 – Access Control
CIA
Confidentiality

assurance that information is not disclosure to
unauthorized programs, users, processes

encryption, logical and physical access control,

The data needs to be classified
Integrity
protecting data or a resource from being altered in an
unauthorized fashion
Availability
fault tolerance and recovery procedures
depends on business and value to business
IAAA
Identification
ensuring that a subject is who he says he is
Unique user name, account number etc OR an issuance
(keycard)
must be non descriptive (you can’t see what someone
can do by the name)
First piece of credentials
Authorization
like password, phrase key token, pin
looking at access control matrix or comparing security
labels
Stacking of authorizations is called Authorization Creep,
too much rights is called excessive privileges
Granted privileges and system granted default access
default no access, give only access that’s needed ( =
NEED TO KNOW)
Second piece of credentials
Strong Authentication if you use 2 out of the three
authentications (know, has, is)AKA 2-factor
authentication
Something a person KNOWS, HAS, IS (knowledge,
ownership, characteristics)
Accountability
each subject is uniquely identified and actions are
recorded
Logical Access Controls: tools used for IAAA
Controls
Primary controls
Administrative
Preventive: hiring policies, screening security
awareness (also called soft-measures!)
Detective: screening behavior, job rotation, review of
audit records
Technical (aka Logical)
Preventive: protocols, encryption, biometrics
smartcards, routers, firewalls
Detective: IDS and automatic generated violation
reports, audit logs
Physical
Preventive: fences, guards, locks
Detective: motion detectors, thermal detectors video
cameras
Operational controls
Detective, Preventive (PASSWORDS TOO), Corrective(restore
controls), Restore control (restore resources) deterrents
Types
Mandatory access control
Authorization depended on security labels which indicate
clearance and classification of objects (Military). Restriction: need
to know can apply. Lattice based is part of it! (A as in
mAndatory!). Rule based access control. Objects are: files,
directories and devices
Discretionary access control
Access through ACL's. Discretionary can also mean: Controlled
access protection (object reuse, protect audit trail). User directed
access control (identity based and hybrid based are also forms of
discretionary) Identity Based AC
Non-discretionary access control
A central authority determines what subjects have access based
on policies. Role based/task based. Also lattice based can be
applied (greatest lower, least upper bounds apply)
Something a user knows
PASSWORDS
cheap and commonly used
password generators
user chooses own (do triviality and policy checking)
One-time password aka dynamic password used only once
Static password Same for each logon
Passphrase easiest to remember. Converted to a virtual
password by the system.
Cognitive password: easy to remember like your mother’s
maiden name
Hacking
access password file
brute force attack (try many different characters) aka
exhaustive
dictionary attack (try many different words)
Social engineering (convince an individual to give
access)
Rainbow Tables (tables with passwords that are already
in hash format
password checker and password hacker
both programs that can find passwords (checker to see if its
compliant, hacker to use it by the hacker)
hashing and encryption
On windows system with utility SYSKEY. The hashed
passwords will be encrypted in their store LM hash and
NT Hash
some OS’s use Seed SALT or NONCE, random values
added to the encryption process to add more complexity
Something a user has
Key, swipe card, access card, badge PASSWORDS.
tokens
Static password token owner authenticates to token, token
authenticates to the information system
Synchronous (TIME BASED) dynamic, uses time or a counter
between the token and the authentication server, secure-ID is an
example
asynchronous (NOT TIME BASED) server sends a nonce
(random value) This goes into token device, encrypts and delivers
a one-time password, with an added PIN its strong authentication
Challenge/response token generates response on a
system/workstation provided challenge
Something a user is
Domain 2 – Access Control
What you do: behavioral
What you are: physical
BIOMETRICS
Most expensive
Acceptable 2 minutes per person for enrollment time
Acceptable 10 people per minute throughput time
IRIS is the same as long as you live
TYPE 1 error: False rejection rate FRR
TYPE 2 error: False Acceptance rate FAR
CER Crossover Error Rate or EER Equal Error rate,
where FRR = FAR. The lower CER/ERR the more
accurate the system.
No sunlight in iris scanner
zephyr chart = iris scans
Finger print: stores full fingerprint (one- to-many
identification), finger scan only the features (one to one
identification).
Finger scan most widely used today
Acceptability Issues: privacy, physical, psychological
T YPES OF B IOMETRICS
 Fingerprints: Are made up of ridge endings and bifurcations
exhibited by the friction ridges and other detailed
characteristics that are called minutiae.
 Retina Scans: Scans the blood-vessel pattern of the retina
on the backside of the eyeball.
 Iris Scans: Scan the colored portion of the eye that
surrounds the pupil.
 Facial Scans: Takes attributes and characteristics like bone
structures, nose ridges, eye widths, forehead sizes and chin
shapes into account.
 Palm Scans: The palm has creases, ridges and grooves
throughout it that are unique to a specific person.
 Hand Geometry: The shape of a person’s hand (the length
and width of the hand and fingers) measures hand geometry.
 Voice Print: Distinguishing differences in people’s speech
sounds and patterns.
 Signature Dynamics: Electrical signals of speed and time
that can be captured when a person writes a signature.
 Keyboard Dynamics: Captures the electrical signals when a
person types a certain phrase.
 Hand Topology: Looks at the size and width of an
individual’s hand and fingers.
Single Sign On (SSO)
Advantage: ability to use stronger passwords, easier
administration, less time to access resources.
Disadvantage: once a key is compromised all resources can be
accessed.
Thin client is also a single sign on approach
Access control methodologies
KERBEROS
Kerberos addresses Confidentiality and integrity and
authentication, not availability
Kerberos Is based on symmetric key cryptology (and is not a
propriety control)
Time synchronization is critical
MIT project Athena
Kerberos is included in windows now (replaced NTLM=NT-LAN
Manager)
Passwords are never exchanged only hashes of passwords
Benefits: inexpensive, loads of OS’s mature protocol
Disadvantage: takes time to administer, can be bottleneck or
single point of failure
The term realm indicates an authentication administrative domain.
Its intention is to establish the boundaries within which an
authentication server has the authority to authenticate a user, host
or service.
Uses symmetric Key cryptography
KDC Key Distribution Center, grants tickets to client for
specific servers. Knows all secret keys of all clients and
servers from the network
AS (Authentication server)
TGS - Ticket granting server
Working:
Client authenticates to the KDC. His passwords becomes an one
way hasted + time = secret key to the AS and gets a TGT Ticket
Granting Ticket,
Client then accesses the TGS with the TGT he has and gets a
ticket to service.
Then the user can use this ticket to service to use the service
SESAME
Public Key Cryptology
European
Needham-Schroeder protocol
Weakness: only authenticates the first block and not the complete
message
Two tickets:
One authentication, like Kerberos
Other defines the access privileges a user has
Works with PACS (Privileged Attribute Certificates)
sesame uses both symmetric as asymmetric encryption
(thus improvement upon Kerberos)
KRYPTOKNIGHT
IBM – thus RACF
Peer-to-peer relationship between KDC and parties
SCRIPTING
scripts contain logon information that authenticates users
DIRECTORY SERVICE
Hierarchical naming schema
active directory has sophisticated security resources (group
policy, user rights accounts, DNS services)
Centralized access control
RADIUS
Remote connectivity via dial in (user dials in to access server,
access server prompt for credentials, user enters credentials and
forwards to radius server, radius server accepts or rejects). USES
UDP. Incorporates an AS and dynamic/static password
DIAMETER= remote connectivity using phone wireless etc, more
secure than radius
CALLBACK; system calls back to specific location (danger in user
forwarding number)
CHAP (part of PPP) supports encryption
TACACS: user-id and static password for network access via TCP
XTACACS separates authentication, authorization and accounting
processes
TACACS+: stronger through use of tokens
Decentralized access control
Databases
Relational databases allow queries
Object oriented databases do not support queries
3 parts
-
Data structures called tables or relations
Integrity rules
Operators on the data in tables
Relation: basis of the database consists of a two dimensional
table
ROWS are records of tuples. Number of rows is cardinality
COLUMNS are attributes. Number of columns is the degree
PRIMARY KEY: unique identifier in a table
Foreign Keys: used to enforce relationship between two tables.
This is also called referentional integrity, that you don’t have a
nonexistent reference.
Smart Cards
IEC 14443 = smartcards
The combi-card -- also known as a dual-interface card -- has one
smart chip embedded in the card that can be accessed through
either contact pads or an embedded antenna.
Smarter than storage cards
Storage smart card holds RSA key pairs in memory
RSA smart cards have processor that compute (sign
and verify RSA certificates) and create RSA key pairs
Identity management
Performs all of IAAA
Directory based
hierarchical x500 standard protocol like LDAP for
allowing subjects to interact with the directory
Organized through name spaces (Through
Distinguished names )
Needs client software to interact
META directory gathers information from multiple
sources and stores them into once central directory and
synchronizes
VIRTUAL directory only points where the data resides
Domain 2 – Access Control
Web Access Management
allows administrators to control what users can access
when browsing enterprise assets
mostly working as stateless HTTP, during session you
are authenticated, once logged of you have to reindentify and authenticate
Can also work as Single Sign on by use of SSL where
through the use of COOKIES the authentication is being
held in memory (preferably) or text file
-
Provisioning
o
user information taken from HR
(authoritative source)
o
Identity data put in an centralized directory
(identity repository)
o
manager will appoint new employees,
accounts are created automatically
o
user provisioning refers to creation,
maintenance and deactivation of user
objects and attributes on systems,
directories or application in response to
business processes.
Profile update
collection of data associated with identity is called a
profile
self service is it called when a user can update his
own non-sensitive data
digital entity is made up of different attributes (like
manager, sex height etc) has clearance level yyy etc
Federation = sharing identity and authentication
behind the scenes (like booking flight --> booking
hotel without re authenticating) by using a federate
identity so used across business boundaries
IPS Intrusion prevention system
Detect attack and PREVENT that attack being successful
Penetration testing
Blue team had knowledge of the organization, can be done frequent
and least expensive
Red team is external and stealth
White box ethical hacker knows what to look for
Black box ethical hacker not knowing what to find
4 stages: planning, discovery, attack, reporting
vulnerabilities exploited: kernel flaws, buffer overflows, symbolic
links, file descriptor attacks
other model: footprint network (information gathering) port scans,
vulnerability mapping, exploitation, report
scanning tools are used in penetration tests
flaw hypotheses methodology = operation system penetration testing
Other things to know
Constrained user interfaces limit the functions that can be selected
by a user
threat: something that could happen to a system,
vulnerability: is a weakness or hole in the security
Password Management
Password Synchronization. Systems synchronize the
passwords to multiple systems. User has one password
but has to re-authenticate at every system. Danger: if
one password is hacked, all resources can be
accessed. Differs from legacy sign on: Users
authenticates once then will gain access without reauthentication
Self-Service password reset. Personal questions (pet’s
name, mother’s maiden name). Often done by question,
then sending mail with link so identity tied to the answer
Assisted password reset. Help Desk authenticates you
by question and answer
Network security
Account management
life cycle management (creating, modifying and deleting
accounts)
Can be automatically or by tickets for technical
administrators on request of the managers
mainly for internal accounts
HOST BASED
monitoring servers through EVENT LOGS AND
SYSTEM LOGS
as good as the completeness of the host logging
Hardening an operation system: disable services and remove
unnecessary applications
Signature based method (AKA Knowledge based): compared
with signature attack database (aka misuse detector)
Statistical anomaly based: defines a ‘normal’ behavior and
detects abnormal behaviors.
Categories within a security label are used to enforce need to know
NIST 800-42 = security testing
War driving: driving a car with notebook to find open access
point to a network
IDS intrusion detection system
NETWORK BASED
Detects intrusions on the local area network behind a
firewall.
Is passive while it acquires data.
Reviews packets and headers
Problem with network based is that it will not detect
attacks by users logged into hosts
Response box is a part of an IDS that initiates alarm or activity
Components: Information source/sensor, centralized monitor
software, data and even report analysis, database components
and response to an event or intrusion
Race Condition: when two or more processes use the same
resource and the sequence of steps within the software can be
carried out in an improper order, thus like force the authorization
step to take place before the authentication step.
TOC/TOU Attack is an asynchronous attack when an attacker
interrupts a task and changes something to affect the result
The system key (SYSKEY) protects security information (including
password information) in the Active Directory database and other
Local Security Authority (LSA) secrets against offline attacks by
encrypting their storage on a domain controller in a Windows server
allowing downloads on a honey pot = illegal (entrapment)
fault generation = getting the encryption key
Network Availability
Raid levels
Domain 3 – Telecommunications and Network
Security
RAID 0 Striped, one large disk out of several –Improved
performance but no fault tolerance
RAID 1 Mirrored drives –fault tolerance from disk errors and single
disk failure, expensive
RAID 2 not used commercially. Hamming Code Parity
RAID 3 Striped on byte level with extra parity drive –Improved
performance and fault tolerance, but parity drive is a single point
of failure and write intensive.
RAID4 Same as Raid 3 but striped on block level
RAID 5 Striped on block level, parity distributed over all drives –
requires all drives but one to be present to operate hotswappable. Interleave parity
RAID 6 Dual Parity, parity distributed over all drives –requires all
drives but two to be present to operate hot- swappable
RAID 7 is as raid5 but all drives act as one single virtual disk
0+1 –striped sets in a mirrored set (minimum four disks; even
number of disks)
Server fault Tolerant Systems
Redundant servers – applies raid 1 mirroring concept to servers.
On error servers can do a fail-over. This AKA server fault
tolerance
Server clustering – group of independent servers with are
managed as a single system. All servers are online and take part
in processing service requests. On error on a server only
performance is affected.AKA server farm
Single point of failures
Cabling
Coaxial many workstations, length.
Twisted pair to long. Cat 5 better than cat3 for interference
Fiber optics immune to EMI, can be broken and highcost/expertise
Topology failures
Ethernet twisted pair more resistant than coaxial
Token Ring because a token is passed by every station, a NIC
that’s is set to wrong speed or error can take all network down
Fiber Distributed Data Interface form of token ring that has second
ring that activates on error
Leased lines use multiple lines and/or multiple vendors
Frame Relay WAN over a public switched network. High Fault
tolerance by relaying fault segments to working.
Network abuse
Class A : unauthorized access by circumventing access controls.
Legitimate users that gain higher access or pretends to be
another user (masquerading)
Class B – unauthorized use of network for non business
properties
Surfing internet, porn sites, private emails
Class C – Eavesdropping
Interception of network traffic. Tapping = physical interception like
clipping
Passive eavesdropping: monitoring or listening to transmissions
Active eavesdropping: tampering with an transmission to create
covert channels or actively probing the network
Class D – Denial of service or other service disruptions (see under
network attacks)
Class E – Network intrusion
Spoofing: giving out incorrect information to deliberately
induce a user or device
Piggy backing: User leaves session open or intruder
notes credentials by looking over shoulder
Back-door attacks: intrusion via dial-up or external
networks
Class F – Probing
Used to gain a road map of the network by using a sniffer. (mostly
in promiscuous mode where all packages are intercepted in clear
text). Manually by using tools like telnet to see what is listening on
a remote sever. Automatic by software programs that do all the
probing and scanning
Network attacks – Denial of Service
Used to overwhelm a targets resources
Filling up hard drive by using huge email attachments or
file transfers
Sends messages to reset targets host subnets masks
Using up all system resources
DOS - performed by sending malformed packets to a system; can
interrupt service or completely deny legitimate users of system
resources
DDOS – botnet, zombie, massive dos attack using multiple
computers
SMURF – ICMP requires three players (attacker, victim and
amplifying network); attacker spoofs packet header to make it
appear that it originated on the victim system with amplifying
network broadcasting the message.
Countermeasures – disable broadcast at border routers; border
routers should not accept packets that originate within network;
restrict ICMP traffic (Hint IC = Its Smurf though spelled wrong)
FRAGGLE – similar to Smurf but uses UDP
Countermeasures – disable broadcast at border routers; border
routers should not accept packets that originate within network;
restrict UDP traffic; employ IDS; apply appropriate patches.
Land Attack - The attack involves sending a spoofed TCP SYN
packet (connection initiation) with the target host's IP address and
an open port as both source and destination.
The reason a LAND attack works is because it causes the
machine to reply to itself continuously.
SYN FLOOD - TCP packets requesting a connection (SYN bit set)
are sent to the target network with a spoofed source address. The
target responds with a SYN-ACK packet, but the spoofed source
never replies. This can quickly overwhelm a system’s resources
while waiting for the half-open connections to time out. This
causes the system to crash or otherwise become unusable.
Counter: sync cookies/proxies, where connections are created
later
Teardrop - The length and fragmentation offset fields of
sequential IP packets are modified, causing the target system to
become confused and crash.
Common Session Hijacking Attacks:
Session hijacking (Spoofing) - IP spoofing involves altering a
TCP packet so that it appears to be coming from a known, trusted
source, thus giving the attacker access to the network.
TCP sequence number attack – intruder tricks target to believe it
is connected to a trusted host and then hijacks the session by
predicting the targets choice of an initial TCP sequence number
Domain 3 – Telecommunications and Network Security
Network layers OSI MODEL
(later succeeded by TCP/IP)
HINT: All People Seems to Need Data Processing
It encapsulates data when going through the layers
Application – layer 7 – C, AU, I, NR
FTP, SMB, TELNET, TFTP, SMTP, HTTP, NNTP, CDP,
GOPHER, SNMP, NDS, AFP, SAP, NCP, SET. Technology:
Gateways. User data
Presentation – layer 6 – C, AU, Encryption
Translations like EBCDIC/ANSI; compression/decompression and
encryption/decryption. Standards like JPEG, TIFF, MID.
Technology: Gateway. Messages
Session -layer 5 -- None
Inter-host communication, simplex, half duplex, full duplex.
Protocols as NSF, SQL, RADIUS, and RPC. Technology:
Gateway
Transport – layer 4 – C, AU, I
End-to-end data transfer services and reliability. Technology:
Gateways. Datagrams
Protocols: TCP, UDP, SSL, SSH-2, SPX, NetBios, ATP
Network – layer 3 – C, AU, I
Path selection and logical addressing. Technology: Virtual circuits
(ATM), routers. Packets
Message routing, error detection and control of node data are
managed. IP, IPSEC, ICMP, BGP, OSPF, RIP, BOOTP, DHCP,
ZIP, DDP, X.25 and IGMP
Data Link – layer 2 - C
This layer deals with addressing physical hardware.
Translates data into bits and formats them into data frames with
destination header and source address. Error detection via
checksums.
LLC: the Logical Link Control Sub layer. Flow control and error
notification
MAC: the Media Access Control layer. Physical addressing.
Concerns frames, logical topologies and MAC-addresses
Protocols: L2F, PPTP, L2TP, PPP, SLIP, ARP, RARP, SLARP,
IARP, SNAP, BAP, CHAP, LCP, LZS, MLP, Frame Relay, Annex
A, Annex D, HDLC, BPDU, LAPD, ISL, MAC, Ethernet, Token
Ring, FDDI
Physical – layer 1 - C
Coverts bits into voltages or light impulses. Hardware and
software drivers are on this level. It sends and receives bits.
Physical topologies: BUS, MESH, STAR, TREE, RING
Network layers TCP/IP Model
Developed by Department of Defense in the 1970s to support the
construction of the internet
HINT: AHIN
Application – layer 4 (Application/Presentation/Session)
Applications and processes that uses the network
Host-to-Host – Layer 3 (Transport)
End-to-end data delivery
Protocols: TCP and UDP
Internet – Layer 2 (corresponds to OSI network layer)
Defines the IP datagram and handles routing of data across
networks
Protocols: IP, ARP, RARP, ICMP
Network access – Layer 1 (Data link, Physical)
Routines for accessing physical networks and the electrical
connection
Network Protocols
Transmission control protocol TCP – reliable, sequences and
works with acknowledgements. Provides a manageable data flow
to avoid congestions overloading and data loss. (like having a
telephone conversation with someone). Connection Oriented.
User datagram protocol UDP – unreliable, scaled down version
of TCP, no error correction, no sequencing. Less overhead. (like
sending a letter to someone). Connectionless.
Internet protocol IP all hosts have an IP address. Each data
packet has an IP address of sender and recipient. Routing in
network is based upon these addresses. Considered unreliable
datagram service because there’s no guarantee that the packet
will be delivered, not even that its delivered only once and no
guarantee that its delivered in the same sequence that its sent
32 bits long, IPv6 is 128 bits long
Address resolution protocol ARP: Used to match an IP address
to a hardware MAC address. ARP sends out broadcast to a
network node to reply with its hardware address. It stores the
address in a dynamic table for the duration of the session, so
ARP requests are only send the first time
Reverse address resolution protocol RARP: When a hardware
address is known but the IP address has to be found. (like an
diskless machine)
Internet control message protocol ICMP: sends messages
between network nodes regarding the health of the network. Also
informs about rerouting incase of errors. Utility PING uses ICMP
messages to check physical connectivity of the network machines
Telnet terminal emulation enables user to access resources on
another machine. Port 23
File Transfer Protocol FTP for file transfers. Cannot execute
remote files as programs. Authentication. Port 20 and 21
Trivial File Transfer Protocol TFTP stripped down, can only
send/receive but not browse directories. No authentication thus
insecure. Port 69
Network File System NFS protocol that supports file sharing
between two different file systems
Simple Mail Transfer protocol SMTP email queuing. Port 25
Line printer daemon LPD for printing and spooling
X Windows graphical user interface
Simple Networking Management Protocol SNMP collection of
network information by polling the devices from a management
station. Sends out alerts –called traps- to an database called
Management Information Bases (MIBs)
Bootstrap Protocol BootP when wireless workstation is on-lined
it sends out a BootP request with its MAC address to get an IP
address and the file from which it should boot. Replaced by DHCP
DHCP: Dynamic Host Configuration Protocol
Security Enhancement Protocols
TELNET: Remote terminal access and Secure Telnet
REMOTE PROCEDURE CALL: Secure remote procedure call
(SRA)
Security Focused Protocols
At application layer of OSI:
Secure Electronic Transaction (SET) authentication for credit
card transactions. Overtaken by SSL
Secure HTTP S-HTTP) encrypting HTTP documents. Also
overtaken by SSL
At Transport layer of OSI:
Secure Shell (SSH-2) Authentication, compression, confidentiality
and integrity.
Uses RSA certificates for authentication and triple DES for
encryption
Secure Socket Layer (SSL) encryption technology to provide
secure transactions like credit card numbers exchange. Two
layered: SSL record protocol and handshake protocol. Same as
SSH it uses symmetric encryption for private connections and
asymmetric or public key cryptography for peer authentication.
Also uses message authentication code for integrity checking.
Simple Key Management for Internet Protocols (SKIP)
provides high availability in encrypted sessions to protect against
crashes. Exchanges keys on a session by session basis.
Firewalls
Domain 3 – Telecommunications and Network
Security
TYPES
Packet filtering firewall AKA screening router
Examines source/destination address, protocol and ports of the
incoming package. Based on ACL’s access can be denied or
accepted. Is considered a first generation of firewall and
operates at Network or Transport layer of OSI
Application level firewall AKA proxy server
While transferring data stream to another network, it masks the
data origin. Second generation firewall operating at Application
layer of OSI
Stateful inspection firewall
All packages are inspected at the Networking layer so it’s faster.
By examining the state and context of the data packages it helps
to track connectionless protocols like UDP and RPC. Third
generation firewall. Analyzed at all OSI Layers.
Dynamic Packet Filtering firewall
Enables modification of the firewall rule. It provides limited support
for UDP by remembering UDP packages across the network.
Fourth generation.
Kernel Proxy Firewalll / Application level Firewall
Runs in windows NT, modular, kernel based, multiplayer session
evaluation. Uses dynamic TCP/IP stacks to inspect network
packages and enforce security policies. Fifth generation
Firewall architecture
Packet filtering routers
Sits between trusted and un-trusted network, sometimes used as
boundary router. Uses ACL’s. Protects against standard generic
external attacks. Has no user authentication, has minimal
auditing.
Screened-Host firewall system
Has both a packet-filter router and a bastion host. Provides both
network layer (package filtering) as application layer (proxy)
server.
Dual homed host firewall
Consists of a host with 2 NIC’s. One connected to trusted, one to
un-trusted. Can thus be used as translator between 2 network
types like Ethernet/token ring. Internal routing capabilities must
not be enabled to make it impossible to circumvent inspection of
data.
Screened-subnet firewalls
Has also defined a De-Militarized Zone (DMZ) : a small network
between trusted an untrusted.
Socks firewall
Every workstation gets some Socks software to reduce overhead
Virtual Private Networks VPN
DATA NETWORK TYPES
A VPN is created by dynamically building a secure
communications link between two nodes using a secret
encapsulation method via network address translation (NAT)
where internal IP addresses are translated to external IP
addresses.
Local Area Network LAN
Limited geographically to e.g. a building. Devices are sharing
resources like printers, email and files. Connected through copper
wire or fiber optics.
CAN: campus area network, multiple building connected to fast
backbone on a campus
MAN: metropolitan network extends over cities
Wide Area network WAN
Connects LANS over a large geographical area
Internet intranet and extranet
Internet is global, intranet local for use within companies and
extranet can be used e.g. by your customers and clients but is not
public.
VPN Protocols
Hint: TP at end for Tunneling Protocols
Point to Point tunneling protocol (PPTP)
Works at data link layer of OSI
Only one single point-to-point connection per session
Point To Point protocol (PPP) for authentication and
tunneling
Dial-up network use
Layer 2 tunneling protocol (L2TP)
Also in data-link layer of OSI
Single point-to-point connection per session
Dial-up network use
Port 115
IPSEC
Operates at Network Layer of OSI
Enables multiple and simultaneous tunnels
Encrypt and authenticate
Build into IPv6
Network-to-network use
VPN Devices
Is hard- or software to create secure tunnels
IP-sec compatible
Encryption via Tunnel mode (entire data package
encrypted) or Transport mode (only datagram
encrypted)
Only works with IP at Network layer of OSI
NON IP-sec compatible
Socks-based proxy servers Used to reach the internal network
from the outside. Also contains strong encryption and
authentication methods
PTP used in windows machines. Multiprotocol, uses PAP or
CHAP
Dial-up VPN’s remote access servers using PPTP commonly
used by ISP’s
Secure Shell SSH2 not strictly a VPN product but opens a secure
encrypted shell session from the internet through a firewall to a
SSH server
DATA NETWORK SIGNALS
Analog signal Infinite wave form, continuous signal, varied by
amplification
Digital signal Saw-tooth form, pulses, on-off only
Asynchronous sends bits of data sequentially. Same speed on
both sides. Modems and dial-up remote access systems
Synchronous very high speed governed by electronic clock
timing signals
LAN Cables
Twisted pair
Shielded (STP) or unshielded (UTP) Cat 3=10BaseT,
Cat5=100BaseT
Coaxial
More EMI resistant. Baseband: only one single channel,
Broadband: multiple signal types like data, video, audio
Fiber Optic
Most expensive, but hard to tap and resistant to EMI
LAN Transmission Protocols
Carrier Sense Multiple Access CSMA for Ethernet.
Workstations send out packet. If it doesn’t get an
acknowledgement it resends
CSMA with Collision Avoidance workstations are attached by
2 coax cables. In one direction only. Wireless 802.11
CSMA with Collision Detection Only one host can send at the
time, using jamming signals for the rest.
Polling Host can only transmit when he polls a secondary to see
if its free
Token-passing Used in token rings Hosts can only transit when
they receive a clear to send token.
LAN Transmission Methods
Unicast Packet is send from single source to single destination
Multicast source packet is copied and send to multiple destinations
Broadcast source packet is copied and send to all nodes
Domain 3 – Telecommunications and Network
Security
LAN Topologies
BUS all transmissions have to travel the full length of the cable
RING Workstations are connected to form a closed loop
STAR nodes are connected to a central LAN device
TREE bus type with multiple branches
MESH all nodes interconnected
LAN Media Access
Ethernet IEEE 802.3 using CSMA with an BUS-topology
Thinnet: 10base2 with coax cables up to 185 meters
Thicknet: 10Base5, coax up to 500 meters
UTP:
10BaseT=10MBps
100baseT=Fast Ethernet =100MBps
1000BaseT=Gigabit Ethernet=1GBps
Ethernet networks were originally designed to work with more
sporadic traffic than token ring networks
ARCnet uses token –passing in a star technology on coax
Token Ring IEEE 802.5 IBM created. All end stations are connected
to a MAU Multi Access Unit. CAU: Controlled Access Units – for
filtering allowed MAC addresses.
Fiber Distributed Data Interface (FDDI) token-passing dual token
ring with fiber optic. Long distances, minimal EMI interference
permits several tokens at the time active
LAN Devices
Repeaters amplify data signals to extend range (physical)
HUBS connect multiple LAN devices into a concentrator. Is actually a
multi-port repeater (physical)
Bridges Forwards data to all other network segments if it’s not on the
local segment. Operates at level 2 (thus no IP-addressing here)
Switches Will only send data to the specific destination address. It’s
actually a multi-port bridge. (Data link)
Routers opens up data packet, reads hardware or network address
and then forwards it to the correct network
Gateway software that acts as access point to another network or
device that translates between different protocols
LAN extenders remote access, multi layer switch that connects
LANs over a WAN
WAN Protocols
Packet switching technologies
Private Circuit technologies
Dedicated line reserved communication, always available
Leased line can be reserved for communications. Type of
dedicated line.
T1 1,5 Mbps through telephone line
T3 44,7 Mbps through telephone line
E1 European 2048 Mbps digital transmission
Serial Line IP (SLIP) TCP/IP over slow interfaces to
communicate with external hosts (Berkley UNIX, windows NT
RAS)
Point to Point protocol (PPP) improvement on slip, adds
login, password and error (by CHAP and PAP) and error
correction. Data link.
Integrated Services Digital Network (ISDN) combination of
digital telephony and data transports. Overtaken by xDSL
xDSL Digital subscriber Line uses telephone to transport
high bandwidth data to remote subscribers
ADSL Asymmetric. More downstream bandwidth up
to 18,000 feet over single copper cable pair
SDSL Symmetric up to 10,000 feet over single
copper cable pair
HDSL High Rate T1 speed over two copper cable
pairs up to 12,000 feet
VDSL Very High speed 13-52MBps down, 1,5-2,3
Mbps upstream over a single copper pair over 1,00
to 4500 feet
X25 defines point-to-point communication between Data terminal
Equipment (DTE) and Data Circuit Terminating Equipment (DCE)
Link Access Procedure-Balanced (LAPB) created for use with
X25, LAPB defines frame types and is capable of retransmitting,
exchanging and acknowledging frames as detecting out-ofsequence or missing frames
Frame Relay High performance WAN protocol designed for use
across ISDN interfaces. Is fast but has no error correction
Switched Multimegabit DATA Service (SMDS) high speed
communication over public switches networks for exchanging
‘bursts of data’ between enterprises
Asynchronous Transfer mode (ATM) very high bandwidth. It
uses 53-byte fixed size cells instead of frames like Ethernet. It can
allocate bandwidth up on demand making it a solution for Busty
applications. Requires fiber optics.
Voice over IP (VOIP) combines many types of data into a single
IP packet. Cost, interoperability and performance wise it’s a major
benefit.
Circuit-switched networks
There must be a dedicated physical circuit path exist during
transmission. The right choice for networks that have to
communicate constantly. Typically for a telephone company
network Voice oriented. Sensitive to loss of connection
Other important WLAN protocols
Synchronous Data Link Control (SDLC) created by IBM for
mainframes to connect to their remote offices. Uses a polling
media access method. Works with dedicated leased lines
permanent up.
Data link layer of OSI model
High-level Data Link Control (HDLC) extension to SDLC also for
mainframes. Uses data encapsulation on synchronous serial links
using frame characters and checksums. Also data link layer
High Speed Serial Interface (HSSI) Defines electrical and
physical interfaces to use for DTE/DCE communications. Physical
layer of OSI
WLAN devices
Message switching networks
Involves the transmission of messages from node-to-node.
Messages are stored on the network until a forwarding path is
available.
Packet-switched networks (PSN or PSDN)
Nodes share bandwidth with each other by sending small data
units called packets. Packets will be send to the other network
and reassembled. Data oriented. Sensitive to loss of data.
More cost effective than circuit switching because it creates
virtual circuits only when they are needed.
Multiplexors device that enables more than one signal to be send
out of one physical circuit
WAN switches multi-port networking devices that are used in
carrier networks. Connect private data over public data by using
digital signals. Data link layer.
Access servers server that provides dial-in and dial-out
connections to the network
Modems transmits data over telephone lines
Channel Service Unit (CSU)/Data service unit (DSU) digital
interface device used to terminate the physical interface on a DTE
device. They connect to the closest telephone company switch in
a central office (CO)
Domain 3 – Telecommunications and Network Security
Remote Access Technologies
Asynchronous Dial-Up Access This is how everyone connects
to the internet. Using a public switched telephone network to
access an ISP
Integrated Serviced Digital Network (ISDN) communication
protocol that permits telephone line to carry data, voice and other
source traffic. Two types: BRI Basic rate interface and Primary
Rate Interface (PRI)
xDSL uses regular telephone lines for high speed digital access
Cable Modems Via single shared coaxial cable, insecure
because of not being filtered or firewalled
Remote Access Security Technologies
Restricted Address incoming calls are only allowed from specific
addresses on an approval list. This authenticates the node, not
the user!
Callback User initiates a connection, supplies identifying code,
and then the system will call back a predetermined telephone
number. Also less useful for travelling users
Caller ID checks incoming telephone number against an approval
list and then uses Callback. Less useful for travelling users.
Remote Node Security Protocols
Password Authenticate Protocol PAP
Provides identification and authentication of the user using static
replayable passwords. No encryption of user-id or password
during communication
Challenge Handshake Authenticate Protocol (CHAP)
non-replayable challenge/response dialog
Remote Access Authentication Systems
Terminal Access Controller Access Control System TACACS
User passwords are administrated in a central database instead of
individual routers. A network device prompts user for a username
and static password then the device queries a TACACS server to
verify the password. TACACSs does not support prompting for
password change or use of dynamic password tokens. Port 49
TACACS+ Enhanced version with use of two factor
authentication, ability to change user password, ability of security
tokens to be resynchronized and better audit trails and session
accounting
Remote Authentication Dial-In User Service RADIUS Often
uses as stepping stone to the more robust TACACS+. Clients
sends their authentication request to a central radius server that
contains all of the user authentication and network ACL’s
RADIUS does not provide two way authentication, therefore it’s
not used for router-to-router authentication. Port 1812. Contains
dynamic password and network service access information
(Network ACLs)
FTP, RLOGIN and TELNET never uses UDP but TCP
Attenuation is decrease in amplitude as a signal propagates along
a transmission medium
SSL session key length is from 40bit to 256 bit
The bridge connects multiple networks at the data link layer, while
router connects multiple networks at the network layer.
Data backups addresses availability, integrity and recovery but not
confidentiality
IP headers contain 32-bit addresses (in IPv4) and 128 in IPv6. In
an Ethernet local area network, however, addresses for attached
devices are 48 bits long.
Wireless
802.11: 1 or 2 mbps, 2.4Ghz, FHSS or DSSS
802.11b: 11 mbps, only DSSS
Things to know
TCPIP Classes
Class A network number values begin at 1 and end at 127
Class B network number values begin at 128 and end at 191
Class C network number values begin at 192 and end at 223
802.11a: 54 mbps, 5 GHz, Orthogonal Frequency Division
802.11g: 20-54mbps, 2.4GHz
802.11e: QoS
ISDN
BRI B-channel 64Kbps, D-channel 16Kbps
PRI B- and D-channels are 64Kbps
802.16: IEEE 802 Broadband Wireless Access (802 WBA)
802.11i: AES, CCMP, 802.1X authentication.
80211 has CSMA/CA as protocol. Can use DSSS and FHSS (ss
stands for spread spectrum)
802.11b uses only DSSS
Before a computer can communicate with the internet, it needs an
IP-address, a default gateway and a subnet mask
To connect multiple LAN segments you can use Bridges,
Switches and Routers
Fast Ethernet 100Base-TX has as characteristics: 100Mbps data
transmission, 1 pairs Cat5 UTP and max segment of 100 meters
(328 feet)
Unsubnetted netmask is shown as /24
Other word for DMZ is screened subnet
802.11n: 100mbps, 2.4GHz
History
Hieroglyphics - sacred carvings
Scytale: wound papyrus around a wooden rod to see message
Substitution character: shifting 3 character (C3) for example in
the one (mono-alphabet) alphabet system
Cipher disks: 2 rotating disks with an alphabet around it
Jefferson disks: 26 disks that cipher text using a alignment bar
Unix uses rot 13, rotate 13 places in the alphabet
Hagelin machine (M-209) mechanical cryptographic machine)
Enigma: poly-alphabetic substitution cipher machine
Techniques
Substitution: like shifting and rotating alphabets, can be broken by
statistical looking at repeating characters or repeats
Transposition (permutation): scrambled orders for example
vertical instead of horizontal
Vernam cipher (one time pad): key of a random set of nonrepeating characters
Information Theory - Claude Elmwood Shannon
Transposition Permutation is used, meaning that letters are
scrambled. The key determines the positions that the characters
are moved to.
Asymmetric key AKA Public Key Encryption
Domain 4 – Cryptography



Sender and receiver have public and private keys.
Public to encrypt a message, private to decrypt
Slower than symmetric, secret key (100 to 1000)
Public Key Algorithms
RSA (Rivest, Shamir and Adleman) works with one way
math with large prime numbers (aka trap door
functions). Can be used for encryption, key exchange
and digital signatures)
Diffie Hellman Key exchange, about exchanging
secret keys over an insecure medium without exposing
the keys
el Gamal, works with discrete logarithms, based on
Diffie Hellmann
DSA Digital Signature Algorithm, the US government
equivalent of the RSA algorithm
ECC - Elliptic Curve Cryptosystem - mathematical
properties of elliptic curves, IT REQUERES FEWER
RESOURCES THAN RSA. Used in low power systems
(mobile phones etc)
Digital signatures
no modifications allowed
identity can be derived
works with a one-way hash (message digest), like SHA-1 (512 bit
blocks) or MD5 (128 bits digest) or HMAC that uses a key
Symmetric key AKA private key or same key
encrypt.




Both the receiver and the sender share a common
secret key
Larger key size is safer > 128
Can be time-stamped (to counter replay attacks)
Does not provide mechanisms for authentication and
non-repudiation
DES (data Encryption Standard) comes from IBM
DEA Data Encryption Algorithm x3.92, using 64 block
size and 56bit key with 8bits parity
16-rounds of substitution and transposition
cryptosystem
Adds confusion(conceals statistical connect between
cipher text and plaintext) and Diffusion (spread the
influence of plaintext characters over many cipher text
characters by means of transposition like HIDE IHED)
Triple des = three times encrypted DES, preferably with
3 different keys = DES-EE3. Actual key length = 168
bits. Uses 48 rounds of computations (3x16)
Replaced by AES Advanced Encryption Standard
4 modes
CBC Cipher Block Chaining - blocks of 64 bits with
64bits initialization vector. Errors will propagate
ECB Electronic Code Book - right block/left block
pairing 1-1. Replication occurs.
Cipher Feedback CFB - stream cipher where the
cipher text is used as feedback into key generation.
errors will propagate
Output Feedback OFB - stream cipher that generates
the key but XOR-ing the plaintext with a key stream. No
errors will propagate
AES Advanced Encryption Standard
Rijndael Block Cipher Algorithm
for speed, simplicity and resistance against known
attacks. Variable block length and variable key lengths
(128,192 and 256 bits)
Not selected for AES were:
RC5 variable algorithm up to 2048 bits key size
IDEA International Data Encryption Algorithm
64 bit plaintext and 128 key length with confusion
and diffusion
used in PGP software
patented requires licenses fees
Two fish key lengths 256 bits blocks of 128 in 16rounds
Blowfish by Bruce Schneider key lengths up to 448 bits
Serpent 32 rounds, 1024 lookups
Hybrid systems
-
-
uses both asymmetrical and symmetrical encryption
o
asymmetrical for key exchange
o
symmetrical for the bulk - thus it is fast
example: SSL, PGP, IPSEC S/MIME
PKI
X.509 standard = PKI
Integrity (hash code and message digest), access control,
confidentiality (by encryption), authentication (digital certificates)
and non-repudiation (digital signatures)
issuer signs a certificate
If you only want to check if a mail is not altered: use digital
signature!
trust anchor = public key that has been verified and that’s trusted
escrowed encryption standard
-
-
legal monitoring of encrypted transmissions
divide key into 2 parts, store separately with two
trusted organizations
part of hardware: US Government Clipper Chip
with Skipjack Secret Key algorithm, but its 80-bits
and weak
Fair Cryptosystems , each portion of a distributed
private key can be verified as correct
email security
S/Mime Confidentiality (encryption) Integrity (using PKCS X.509
PKI) and non-rep through signed message digests
PEM Privacy Enhanced Email Encryption (AES) PKI X.509 and
RSA.
Message Security protocol Military X.400. Sign, Encrypt, Hash
Pretty Good Privacy, uses IDEA and RSA instead of an CA they
use Web of Trust between the people using it
Internet Security
Definitions
Things to know
Message Authentication Code MAC or Financial Institution
Message Authentication Standard FIMAS
Macs checks value like CRC
Purpose: protect transmitted information from being read and
understood expect the intended recipient
skip is a distribution protocol
SET Secure Electronic Transaction
Uses DES for encrypting payment information for credit card
companies. Covers end to end transactions with digital signatures
and digital certificates
SSL / TLS
Netscape!
Operates at the TRANSPORT layer
des, idea, 3des and md5 hash
x.509 PKI certificates
does not offer end to end transactions
based on sessions
40 or 128 bits
HTTPS running http over SSL, encrypts message and connection
SHTTP protects only individual message between computers
secure Shell SSH-2 remote access to network using encrypted
tunnels
Domain 4 – Cryptography
WAP
class1: anonymous authentication
class2: server authentication
class3: two way client server authentication
WEP: initialization vector, RC4 and symmetric = old do not use
IOTP Internet Open trading Protocol
user chooses payment method and thus encryption
MONDEX cash smart card application
IPSEC
2 protocols: AH Authentication header and ESP Encapsulated
Security Payload
works with Security Associations (SA's)
works with IKE protocols IKE IS FOR MANAGING SECURITY
ASSOCIATIONS
2 modes:
transport, data is encrypted header is not
tunneled: new uses rc6; IP header is added, old IP header and
data is encrypted
Block Cipher: segregating plaintext into blocks and applying
identical encryption algorithm and key
Cipher: cryptographically transformation that operates on
characters or bits
Cipher text or Cryptogram: unintelligible message
Clustering: situation wherein plain text messages generates
identical cipher text messages using the same algorithm but with
different crypto-variables or keys
Codes: cryptographic transformation that operates at the level of
words or phrases
Cryptanalysis: breaking the cipher text,
Cryptographic Algorithm: Step by step procedure to encipher
plaintext and decipher cipher text
Cryptography: the art and science of hiding the meaning of
communications from unintended recipients. (Greek:
kryptos=hidden, graphein=to write)
Cryptology: cryptography + cryptanalysis
Cryptosystem: set of transformations from a message space to
cipher space
Decipher: To make the message readable, undo encipherment
process
Encipher: make message unintelligible
End-to-end encryption: Encrypted information that is sent from
point of origin to destination. In symmetric encryption this means
both having the same identical key for the session
Exclusive OR: Boolean operation that performs binary addition
Key or Crypto variable: Information or sequence that controls
the enciphering and deciphering of messages
Link encryption: stacked encryption using different keys to
encrypt each time
One Time Pad: encipher each character with its own unique key
that is used only once
Plaintext: message in clear text readable form
Steganography: secret communications where the existence of a
message is hidden (inside images for example)
Work Function (factor): the difficulty of obtaining the clear text
form the cipher text as measured by cost/time
RC4 is a stream cipher
rc5 and rc6 are block cipher
FIPS 140 hardware and software requirements
Hash algorithms are also called Message Digests.
Most used are MD5 (message Digest 128 bits) and SHA1
(signature hashing algorithm 160 bits)
CRL’s of a PKI environment holds serial numbers
SHA1 was designed by NIST and NSA to be used in digital
signatures
A root Certificate Authority (CA) must certify its own public key
pair
cross certification does not check authenticity off the certificates in
the certificates path
Traffic analysis: inference of information from analysis of traffic
Traffic padding: generation of spurious data units
Collision: Same message digest as a result of hashing.
Cryptographic Attacks:
Ciphertext Only: attacker sees only the ciphertext
Known Plaintext: attacker knowns both cipher and plaintext
Chosen Plaintext: offline attack (attacker prepares list of
plaintexts) online attack (attacker chooses the plaintext based on
the ciphertext already received)
Chosen ciphertext: attacker chooses both the plaintext values and
the ciphertext values
3 states of information
data at rest (storage)
data in transit (the network)
data being processed (must be decrypted)
cipher types: block (padding to blocks of fixed size) like DES
3DES AES or stream (bit/byte one by one no padding) like RC4
Chameleon, leviathan, Sober
Birthday Attack: Collisions appear much fasters.
Domain 5 – Security Architecture and Models
Computer Architecture
Primary Storage is a temporary storage area for data entering
and leaving the CPU
Random Access Memory (RAM) is a temporary holding place for
data used by the operating systems. It is volatile; meaning if it is
turned off the data will be lost. Two types of RAM are dynamic
and static. Dynamic RAM needs to be refreshed from time to time
or the data will be lost. Static RAM does not need to be refreshed.
Read-Only Memory (ROM) is non-volatile, which means when a
computer is turned off the data is not lost; for the most part ROM
cannot be altered. ROM is sometimes referred to as firmware.
Erasable and Programmable Read-Only Memory (EPROM) is
non-volatile like ROM, however EPROM can be altered.
Process states: Stopped, waiting, running, ready
Multitasking: execute more than one task at the same time
Multiprocessing more than one CPU is involved.
Multi Threading: execute different parts of a program
simultaneously
Single state machine operates in the security environment at the
highest level of classification of the information within the
computer. In other words, all users on that system must have
clearance to access the info on that system.
Multi-state machine can offer several security levels without risk
of compromising the system’s integrity.
Protection mechanisms
Recovery procedures
Protection domain
Execution and memory space assigned to each process
Recovery procedures: system should restart in secure mode
Startup should occur in maintenance mode that permits access
only by privileged users from privileged terminals
Fault-tolerant continues to function despite failure
Fail safe system, program execution is terminated and system
protected from compromise when hardware or software failure
occurs
Fail soft or resilient system, selected, non-critical processing is
terminated when failure occurs
Failover, switches to hot backup.
TRUSTED COMPUTER BASE
Combination of protection systems within a computer system,
which include the hardware, software and firmware that are
trusted to enforce the security policy.
Security Kernel is hardware, software, firmware, elements of
TCB that implement the reference monitor concept — must be
isolated from reference monitor (reference monitor: isolation,
completeness and verifiability, that compares the security labels of
subjects and objects)
Protection rings (MIT’s MULTICS design)
Ring 0 - Operating system kernel. The OS’ core. The kernel
manages the hardware (for example, processor cycles and
memory) and supplies fundamental services that the hardware
does not provide.
Ring 1 - Remaining parts of the operating system
Ring 2 - I/O drivers and utilities
Ring 3 - Applications and programs
Security Modes (used in MAC)
CICS complex instructions. Many operations per instruction. Less
number of fetches
RISC reduced instructions. Simpler operations per instruction.
More fetches.
Software
1 GL: machine language (used directly by a computer)
2GL: assembler
3GL: FORTRAN. basic pl/1 and C++
4GL: Natural / focus and SQL
5GL: Prolog, lisp artificial intelligence languages based on logic
Certification and accreditation
Certification is evaluation of security features and safeguards if it
meets requirements
DITSCAP
US defense and government certification
Definition (Phase 1), Verification (Phase 2), Validation (Phase 3),
Post Accreditation (Phase 4)
AND NIACAP
National security certification
Keyword: lifecycle
Offers: Site, Type and System accreditation
Accreditation is the formally acceptance of outcome of
evaluation by management
Dedicated security mode :
All users can access all data.
Clearance for all information.
Need to know for ALL data
system high security mode:
All users can access some data, based on need to
know
Clearance for all information
Need to know for SOME data
compartmented security mode:
All users can access some data, based on their need
to know and approval.
Clearance for all information they access
Need to know for SOME data
Use of information labels
Multi level:
All users can access some data, based on their need
to know, approval and clearance.
Clearance for all information they access
Need to know for SOME data
Others:
controlled type of multilevel security where a limited amount of
trust is placed in the system’s hardware/software along with
classification
limited access: minimum user clearance is not cleared and the
maximum data classification is unclassified but sensitive
Assurance
Degree of confidence in satisfaction of security requirements
Evaluation criteria
Trusted Computer System Evaluation Criteria
TCSEC: (Orange book) From the U.S. DoD, it evaluates operating
systems, application and systems. It doesn’t touch the network
part. It only addresses confidentiality!
-
-
-
D minimal protection, any systems that fails higher
levels
C1 Discretionary protection (identification,
authentication, resource protection).
C2 – AND Controlled access protection (object reuse,
protect audit trail).
B1 Mandatory protection (security labels) based on BellLaPadula security model. Labeled security (process
isolation, devices labels).
B2 AND Structured protection (trusted path, covert
channel analysis). Separate operator/admin roles.
Configration management
B3 AND security domain (trusted recovery, Monitor
event and notification).
A1 – verified design
A – verified protection
Operational assurance requirements for TCSEC are:
System Architecture
System Integrity
Covert Channel analysis
Trusted Facility Management
Trusted recovery
Domain 5 – Security Architecture and Models
Rainbow series:
Red = trusted network,
Orange = TCSEC evaluation
Brown =trusted facilities management,
Tan=audit,
Aqua=glossary.
Green = password management
Information Technology Security Evaluation Criteria
ITSEC: it is used in Europe only, not USA. Addresses CIA.
Unlike TCSEC it evaluates functionality and assurance
separately. Assurance from E0 to E6 (highest) and F1 to F10
(highest). Therefore a system can provide low assurance and
high functionality or vice-versa.
Common Criteria ISO 15408
Defines a protection profile that specifies the security
requirements and protections of a product that is to be
evaluated. Organized around TCB entities. Evaluation
Assurance Levels (EAL)
•EAL0 –Inadequate assurance
•EAL1 –Functionally tested
•EAL2 –Structurally tested
•EAL3 –Methodically tested and checked
•EAL4 –Methodically designed, tested and reviewed
•EAL5 –Semi formally designed and tested
•EAL6 –Semi formally verified design and tested
•EAL7 –Formally verified design and tested
Target of Evaluation (TOE): the product
Protection Profile (PP): Security requirements for a class of
security devices
Security Target (ST): identifies the security properties of TOE
Security Functional Requirements (SFRs): Specific individual
security functions
Models
MATRIX
Provides access rights to subjects for objects
Access rights are read, write and execute
Columns are ACL’s
Rows are capability lists
Supports discretionary access control
TAKE-GRANT
uses a direct graph to specify the rights that subjects can
transfer to objects or that subjects can take from other
subjects
Uses STATES and STATE TRANSTIONS
BELL-LAPADULA
Confidentiality model
developed by DOD, thus classification
Cannot read up (simple e=read security rule)
Cannot write down (* property rule AKA CONFINEMENT
PROPERTY). Exception is a trusted subject.
Uses access matrix to specify discretionary access control
Use need to know principle
Strong star rule: read and write capabilities at the same
level
First mathematical model defined
tranquility principle in Bell-LaPadula prevents security level
of subjects from being changed once they are created
BIBA
-
Integrity model
Cannot read down (simple e=read integrity rule)
cannot write up (* integrity)
lattice based (least upper bound, greatest lower bound,
flow policy)
subject at one level of integrity cant invoke subject at a
higher level of integrity
CLARK WILSON
integrity model
Cannot be tampered, logged, and consistency
Enforces segregation of duty
Requires auditing
Commercial use
Works with SCI Constrained Data items, data item whose
integrity is to be preserved
Access to objects only through programs
Information flow model
Each object is assigned a security class and value, and
information is constrained to flow in the directions that are
permitted by the security policy. Thus flow off information from one
security level to another.
Covert channels
Is a way to receive information in an unauthorized manner.
Information flood that is not protected by a security mechanism.
2 types
Storage covert channel: processes communicate via storage
space on the system
Covert timing channel: one process relays to another by
modulating its use of system resources.
Countermeasures: eal6 systems have less than eal3 systems
because covert channels are normally a flaw in design.
Non interference model
Groups of users are separated with their commands. Ensures that
activities performed at a higher security level do not affect the
activities at a lower security level
Brewer and Nash
The Chinese Wall model provides a dynamic access control
depending on user’s previous actions. This model prevents conflict
of interests from members of the same organization to look at
information that creates a conflict of another member of that
organization. Ex. Lawyers in a law firm with client oppositional
Other things to know
Objects of sensitivity labels are: single classification and
component set
Trusted recovery is: after failure or crash system is still secure
‘dominate’ in access control means access to higher or equal
access class
Security perimeter = line between TCB and outside
Validating TCB = formal for system integrity
Tempest: shielding and other emanations-reducing mechanism
Domain 6 – Operational Security
Categories of Controls
Trusted recovery
Preventive lower the amount and impact of unintended errors
and prevent unauthorized intruders to access the systems
Detective used to detect an error once it has occurred,
operate after the fact. E.g. audit trail
Corrective implemented to help mitigate the impact of a loss
e.g. restoring data
Deterrent controls used to encourage compliance. e.g
Directive controls
Application controls minimize and detect the software’s
operational irregulatrities.
Transaction Controls initiation to output through testing and
change control.
Input controls input must be properly and valid.
E.g. time stamping and counting
Processing controls transactions have to be valid
and improper transactions have to be dealt with
Output controls protecting confidentiality and verify
integrity by comparing with input data
Change controls preserve data integrity while
changes are made
Test controls during testing confidentiality has to be
protected thus sanitized data has to be used
Ensures that the security is not breached when a system crash or
failure occurs. Only required for a B3 and A1 level systems.
Failure preparation Backup critical information thus enabling data
recovery
Administrative Management controls
Separation of duties assigns parts of tasks to different
individuals thus no single person has total control of the
system’s security mechanisms
Least privilege a system’s user should have the lowest level
of rights and privileges necessary to perform their work and
should only have them for the shortest time. Three types:
Read only, Read/write and Access/change
Two-man control two persons review and approve the work
of each other
Dual control two persons are needed to complete a task
Rotation of duties limiting the amount of time a person is
assigned to perform a security related task before being
moved to different task to prevent fraud
Mandatory vacations prevent fraud and allowing
investigations
Need to know the subject is given only the amount of
information required to perform an assigned task
Employment screening or background checks
Violation Analysis
Clipping levels must be established to be effective
Clipping Level – baseline of normal activity, used to ignore
normal user errors
Profile Based Anomaly Detection
Looking for:
Repetitive Mistakes
Individuals who exceed authority
Too many people with unrestricted access
Patterns indication serious intrusion attempts
System recovery after a system crash
1. Rebooting system in single user mode or recovery
console, so no user access is enabled
2. Recovering all file systems that were active during failure
3. Restoring missing or damaged files
4. Recovering the required security characteristic, such as
file security labels
5. Checking security-critical files such as system password
file
Common criteria hierarchical recovery types
1. Manual System administrator intervention is required to
return the system to a secure state
2. Automatic Recovery to an secure state is automatic when
resolving a single failure (though system administrators
are needed to resolve additional failures)
3. Automatic without Undo Loss Higher level of recovery
defining prevention against the undue loss of protected
objects
Threats and vulnerabilities
Accidental loss occurs unintentionally
Examples: user input errors deletion errors faulty data or
application programs
Inappropriate activities computer behavior that does not rise the
level of criminal activity, but may y be grounds for job action or
dismissal
Inappropriate content using corporate property to store
illegal content as porno, entertainment political data
Waste of corporate resources using corporate
resources for private use
- Sexual or Racial Harassment using computer
resources
- Abuse of Privileges and Rights
Illegal Computer Operations
Eavesdropping – sniffing, dumpster diving, social
engineering
Fraud – collusion, falsified transactions
Theft – information or trade secrets, physical hardware
and software theft
Sabotage – Denial of Service (DoS), production delays
External Attacks – malicious cracking, scanning, war
dialing
Other things to know
OPSEC process- Understanding your day-to-day operations from
the viewpoint of a competitor, enemy, or hacker and then
developing and applying countermeasures.
Pen-test – testing of network security as would a hacker do to find
vulnerabilities. Always get management approvement firs!
Types of system failure
System reboot System shuts itself down in a controlled manner
after detecting inconsistent data structures or runs out of resources
Emergency restart when a system restarts after a failure happens
in an uncontrolled manner. E.g. when a low privileged user tries to
access restricted memory segments
System cold start when an unexpected kernel or media failure
happens and the regular recovery procedure cannot recover the
system in a more consistent state.
Port scanner: program that attempts to determine whether any of a
range of ports is open on a particular computer or device
Ring zero- inner code of the operating system. Reserved for
privileged instructions by the OS itself
War dialer: dials a range of phone numbers as in the movie wargames
Assurance = other word for security
Monitoring and auditing
Companies can set predefined thresholds for the number of certain
types of errors that will be allowed before the activity is considered
suspicious. This baseline is referred to as clipping level
Audit trails
-
Transaction date/time
Who processed the transaction
At witch terminal
Various security events
Superzapping: system utility or application that bypasses all
access controls and audit/logging functions to make updates to
code or data
Operational assurance – Verification that a system is operating
according to its security requirements

Design & development reviews

Formal modeling

Security architecture

ISO 9000 quality techniques

Assurance – degree of confidence that the implemented
security measures work as intended
Piggybacking: when an unauthorized person goes through a door
behind an authorized person.
Supervisor mode: processes running in inner protected ring
Operational Assurance
Operational Assurance – focuses on basic features and
architecture of a system

System Architecture

System Integrity

Covert Channel Analysis

Trusted Facility Management

Trusted Recovery
Configuration Change Management – Required B2, B3 and A1







Covert Channel Analysis
An information path that is not normally within a system and is
therefore not protected by the systems’ normal security mechanism.
Secret ways to convey information to another program or person

Covert Storage Channels - convey information by
changing stored data (B2)


Domain 6 – Operational Security

Covert Timing Channels – convey information by
altering the performance of or modifying the timing of
system resources in measurable way. (B3, A1= Storage
and Timing)
Combat Covert Channel Analysis - with noise and traffic
generation
Trusted Facility Management - Required for B2, B3, and A1
Defined as assignment of a specific individual to administer the
security of a system. (Security Administrator)


Separation of Duties
B2 security level requires that systems must support
separate operator and system administrator roles.
B3 and A1, systems must clearly identify the
functions of the security administrator to perform the
security-related functions.
Rotation of duties
Trusted Recovery - Required for B3 and A1 levels
Life Cycle Assurance
Life Cycle Assurance – controls and standards required for
building and maintaining a system

Security Testing

Design Specification and testing

Configuration Management

Trusted Distribution
Process of tracking and approving changes
Identify, control and audit changes
Changes to the system must not diminish security
Includes roll back procedures
Documentation updates to reflect changes
Recommended for systems below the required B2, B3 and
A1
Change Control Functions:
Orderly manner and formalized testing
Users informed of changes
Analyze effects of changes
Reduce negative impact of changes
Configuration Management required for Development and
Implementation stages for B2 and B3
Configuration Management required for life cycle of
system for A1
Operations Controls
Resource Protection
Protecting Resources from disclosure alteration or misuse
Hardware – routers, firewalls, computers, printers
Software – libraries, vendor software, OS software
Data Resource – backup data, user data, logs
Hardware Controls
Hardware Maintenance
Requires physical and logical access by support and vendors /
Supervision of vendors and maintenance, background checks
Maintenance Accounts
Disable maintenance accounts when not needed
Rename default passwords
Diagnostic Port Control
Specific ports for maintenance
Should be blocked from external access
Hardware Physical Controls – require locks and alarms
Sensitive operator terminals
Media storage rooms
Server and communications equipment
Modem pools and circuit rooms
Software Controls
Anti-virus Management – prevent download of viruses
Software Testing – formal rigid software testing process
Software Utilities – control of powerful utilities
Safe software Storage – prevent modification of software and copies
of backups
Back up Controls – test and restore backups
Privileged Entity Controls –“ privileged operations functions”
Extended special access to system commands
Access to special parameters
Access to system control program – some only run in particular state
Media Resource Protection
Media Security Controls – prevent the loss of sensitive information
when the media is stored outside the system
Logging – log the use of the media, provides
accountability
Access Control – physical access control
Proper Disposal – sanitization of data – rewriting,
degaussing, destruction
Media Viability Controls – protect during handling, shipping and
storage
Marking – label and mark media, bar codes
Handling – physical protection of data
Storage – security and environmental protection from
heat, humidity, liquids, dust, smoke, magnetism
Physical Protection
Protection from physical access
Hardware – routers, firewalls, computers, printers
Software – libraries, vendor software, OS software
Physical piggybacking – following an authorized person through a
door
Penetration Testing
Testing a networks defenses by using the same techniques as
external intruders





Scanning and Probing – port scanners
Demon Dialing – war dialing for modems
Sniffing – capture data packets
Dumpster Diving – searching paper disposal areas
Social Engineering – most common, get information by
asking
Problem Management
Goals of



problem management:
Reduce failures to a manageable level
Prevent occurrence of a problem
Mitigate the impact of problems
Potential Problems:

Performance and availability of computing resources

The system and networking infrastructure

Procedures and transactions

Safety and security of personnel
Abnormal Events - that can be discovered by an audit

Degraded resource availability

Deviations from the standard transaction procedures

Unexplained occurrences in a processing chain
Objective of problem management is resolution of the problem
Domain 7 – Applications and Systems development
Software Life Cycle Development
MODELS
Simplistic model
This model was simplistic in that it assumed that each step could
be completed and finalized without any effect from the later
stages that may require rework.
Waterfall model
Can be managed if developers are limited going back only one
step. If rework may be done at any stage it’s not manageable.
Problem: it assumes that a phase or stage ends at a specific time.
System Requirements-> Software Requirements -> Analysis ->
Program Design -> Coding -> Testing -> Operations &
Maintenance
Waterfall including Validation and Verification (V&V)
Reinterpretation of the waterfall model where verification
evaluates the product during development against specification
and validation refers to the work product satisfying the real-world
requirements and concepts.
Verification=doing the job right
Validation:= doing the right job
Spiral model
Angular = progress made
Radial = cost
Lower left = development plans
Upper left = objectives of the plans, alternatives checked
Upper right = assessing alternatives, risk analysis
Lower right = final development
Left horizontal axis = includes the major review required to
complete each full cycle
LIFECYCLE
Information security in Lifecycle Management
Conception phase: Policies, standards, threats vulnerabilities
legal, cost etc.
Initiation phase: think about encryption and security specs
Development phase: incorporate security specs. Determine
access controls, verification
Implementation phase: install security software
Testing phase: test security software and controls, documentation
Maintenance phase: revalidate controls pen test, change process
Testing issues
Personnel separate from developers should test.
Should also check for incorrect data types and data out of range
(live of actual data might not do so)
UNIT TESTING; testing small piece of software during a
development stage by developers and quality assurance
Maintenance and change control
Request control: manages users requests, sets priority, costs
and interface
Change control: recreating and analyzing the problem,
developing the change, quality control, tools, documentation,
restrictions and recertification and accreditation if necessary
Release control: issuing latest release of software
Configuration management
Configuration item (CI) component whose state is recorded
Version: recorded state of the CI
Configuration: collection of component CI’s that make another CI
Building: assembling a version of a CI using component CI’s
Build list: set of versions of component CI’s used to build a CI
Software Library: controlled area only accessible for approved
users
Software capability maturity model (CMM)
Quality of software is a direct function of quality of development
and maintenance
Defined by Carnegie Mellon University SEI (Software Engineering
Institute)
Describes procedures, principles, and practices that underlie
software development process maturity
5 levels
1- initiating – competent people, informal processes ad-hoc
2- repeatable – project management processes
3 - defined – engineering processes
4 – managed – product and process improvement, quantitatively
controlled
5 – Optimizing – continuous process improvement
Works with an IDEAL model.
Initiate begin effort, Diagnose perform assessment, Establish an
action plan, Action implement improvements, Leverage
reassesses and continuously improve
Object-Orientated systems
Objects behave as black box; they are encapsulated to perform
an action. Can be substituted if they have compatible operations.
It can store objects like video and pictures
Message: communication to object to perform an action
Method: code that defines an action an object performs in
response to a message
Behavior: results exhibited by an object in response to a message
Class: collection of methods that defines the behavior of objects
Instance: objects are instances of classes that contain their
methods
Inheritance: methods from a class are by subclass
Multiple Inheritance: class inherits characteristics from more than
one parent class
Delegation: forwarding a request to another object
Polymorphism: objects of many different classes that are related
by some common super class.
Poly-instantiation: development of detailed version of an object
from another object using different values in the new object
5 phases of object orientation
Requirements analysis (OORA) defines classes of objects and
their interactions
Analysis (OOA) understanding and modeling a particular problem
Domain Analysis (DA) seeks to identify classes and objects that
are common to all applications in a domain
Design (OOD)Objects are the basic units, and instances of
classes
Programming (OOP) employment of objects and methods
If class = airplane, objects like fighter plane, cargo plane,
passenger plane can be created. Method would be what a plane
would do with a message like: climb, dive, and roll.
Object Request Brokers (ORBs): middleware that acts as locators
and distributors of the objects across networks.
Standards
Common object request broker (CORBA) architecture enables
programs written in different languages and using different
platforms and OS’s through IDL (Interface Definition Language)
Common object Model (COM) support exchange of objects
amongst programs. This used to be called OLE. DCOM is the
network variant (distributed)
Conclusion: Object orientation (e.g. with C++ and Smalltalk)
supports reuse of objects and reduces development risk, natural
in its representation of real world entities.
Cohesion: ability to perform without use of other programs
High cohesion: without use of other modules
Low cohesion: Must interact with other modules
Coupling: affect on other modules.
High coupling: module largely affects many more modules
Low coupling: it doesn’t affect many other
Security Life Cycle Components
System feasibility: ISP, Standards, Legal, validation concepts
Software Plans & Requirements: Due diligence, Threats,
Security requirements
Product Design: incorporate security specs, Design docs,
Determine access controls, Verification
Detailed Design: Design access controls & security controls,
detailed docs, verification, Consider BCP, employ encryption
Coding: Unit testing, Support BCP, develop docs
Integration product: Refine docs, integrate security, security
verification, test integrated modules
Implementation: Install, Test security, Run system, Acceptance
testing, Complete documentation, certification, and accreditation
Operations & Maintenance: Pen test, Change control, update
docs, recertification, revalidate security controls
Domain 7 – Applications and Systems development
Artificial intelligence systems
Database Security Issues
Real-time systems
Expert Systems
Based on human reasoning
Knowledge base of the domain in the form of rules
If-then statements=called forward chaining
Priority in rules are called salience
Interference system = decision program
Expert system = inference engine + knowledge base
Degree of uncertainty handled by approaches as
Bayesian networks(probability of events), certainty
factors(probability an event is true) or fuzzy logic(to
develop conclusions)
Two modes:
o
Forward chaining: acquires info and comes to
a conclusion
o
Backward chaining: backtracks to determine
IF a hypothesis is correct
Neural Networks
Based on function of biologic neurons
Works with weighted inputs
If a threshold is exceeded there will be output
Single-layer : only one level of summoning codes
Multi-level: more levels of summoning codes
Training period needed to determine input vectors
adaptability (learning process)
Security can be provided through Views. That is a virtual relation
that combines information from other relations. A view can be
used to restrict data made available for users based on their
privileges and need-to-know.
operations: join, project and select (JPS)
Views will hide information that a user is not allowed to see, thus
implementing the LEAST Privilege.
Acquire data from transducers or sensors in real time, and then
making computations and control decisions in a fixed time
window. (e.g. fly-by-wire on airplanes)
Availability is crucial; hence RAID systems are in place.
Granularity is the fineness in with access can be controlled or
limited
OLTP Online Transaction Processing: clustered databases to
provide fault tolerance and high performance. Insure that
transactions happen properly or not at all.
Transactions sometimes described as ideally ACID
Atomic: divides transactions into units of work, all modifications
take effect or none ( then a rollback)
Consistent: all records follow integrity constraints
Isolated: transactions executed in isolation until completed
Durable: not reversible once committed
Aggregation is the act of obtaining information of a higher
sensitivity by combining information of lower levels of sensitivity.
Inference: use the ability of users to deduce information about
data at unauthorized levels using inference channels
Data warehousing
Data warehousing is a repository of information from
heterogeneous databases that is available for users for making
queries. Data is normalized (Ensures that attributes in a table only
depend on the primary key)
Data mining
Database systems
Database: general mechanism for defining, storing and
manipulating data without writing specific programs
DBMS: refers to a suite of software programs that maintains and
provides controlled access to data components store in rows and
columns of a table
Types
-
Data mining is searching for data correlations in the data
warehousing.
The correlation of data about data is called metadata.
Can be stored in a separate database with high levels of
protection called the Data mart. The information obtained from
data marts can be send back to the data warehouse
Data dictionaries
Hierarchical= tree (sons with only one parent)
Network = tree (all interconnected)
Mesh
Object-orientated
Relational –has DDL and DML, has TUPLES and
ATTRIBUTES (rows and columns)
DDL – Data definition language defines structure and schema
DML – Data manipulation language view, manipulate and use
the database via VIEW, ADD, MODIFY, SORT and DELETE
commands.
DDE – Dynamic data exchange enables applications to work in a
client/server model by providing the interprocess communications
mechanism (IPC)
DCL – Data control language subset of SQL used to control
access to data in a database, using GRANT and REVOKE
statements
Fault tolerant: the system has to detect a fault and take action to
recover
Data dictionary is a database for system developers. It records all
data structures used by an application. If a data dictionary is
separated, the primary dictionary provides the baseline of the data
and the central control, the secondary dictionary to separate
development projects, provide backup for primary and to serve as
a partition between the development and test databases
Centralized Architecture
Centralized systems are less difficult to protect because they are
not interconnected through a network
Distributed systems are interconnected through a network.
Things to know
Noise and perturbatation: inserting bogus information to hope to
mislead an attacker
CASE= tool for development
First step by change process = management approval.
NB: when a question is about processes, there must always be
management’s approval as First step.
PROTOTYPING: customer view taken into account
SQL -SUDIGR
Select, Update, Delete, Insert, Grant, Revoke
Bind variables are placeholders for literal values in SQL query
being sent to the database on a server
Bind variables in SQL used to enhance performance of a
database
Monitor progress and planning of projects through GANTT and
PERT charts
Check digit: point of verification in an computerized application
Semantic integrity: make sure that the structural and semantic
rules are enforced on all data types, logical values that could
adversely affect the structure of the database
Referential integrity: all foreign keys reference existing primary
keys
Foreign and primary keys uniquely identify a record in a
database
Dynamic Lifetime Objects: Objects created on the fly by
software in an Object Oriented Programming environment.
An object is preassembled code that is a self-contained module
Domain 7 – Applications and Systems development
Mobile code
Virus
System Development Life Cycle
Java – sandboxes, no warnings, programs are compiled to bytecode
ActiveX – Authenticode, relies on digital signatures, annoying
dialogs people click away
Boot sector – moves or overwrites the boot sector with the
virus code.
Project initiation: Feasibility, cost, risk analysis, Management approval,
basic security objectives
Functional analysis and planning: Define need, requirements, review
proposed security controls
System design specifications: Develop detailed design specs, Review
support documentation, Examine security controls
Software development: Programmers develop code. Unit testing Check
modules. Prototyping, Verification, Validation
Acceptance testing and implementation: Separation of duties,
security testing, data validation, bounds checking, certification,
accreditation
Operations and maintenance: release into production.
Certification/accreditation
Revisions/ Disposal: remove. Sanitation and destruction of unneeded
data
Malicious code threats
Virus reproduces using a host application. It inserts or attaches
itself to the file
Worm reproduces on its own without host application
Logic Bomb/Code Bomb executes when a certain event
happens (like accessing a bank account) or a data/time occurs
Trojan Horse program disguised as a useful program/tool
HOAXES – False warnings like: DON’T OPEN X SEND TO ALL
YOUR COLLEGUES
Remote Access Trojan (RAT) remote control programs that
have the malicious code and allow for unauthorized remote
access Back orifice, sub seven, net bus )
Botnet compromise thousands of systems with zombie codes
can be used in DDOS attacks or spammers
Buffer Overflow Excessive information provided to a memory
buffer without appropriate bounds checking which can result in
an elevation of privilege. If executable code is loaded into the
overflow, it will be run as if it were the program.
Buffer overflows can be detected by disassembling programs
and looking at their operations.
Buffer overflows must be corrected by the programmer or by
directly patching system memory.
Trap Door An undocumented access path through a system.
This typically bypasses the normal security mechanisms and is
to plant any of the malicious code forms.
Backdoor program installed by an attacker to enable him to
come back on a later date without going through the proper
authorization channels
Covert Channel Is a way to receive information in an
unauthorized manner. Information flood that is not protected by a
security mechanism.
Covert Storage Channel Writing to storage by one process and
reading by another of lower security level.
Covert Timing Channel One process relays to another by
modulating its use of system resources.
Countermeasures: EAL6 systems have less than EAL3 systems
because covert channels are normally a flaw in design.
LOKI is a tool used for covert channel that writes data directly
after the ICMP header
System infector – infects BIOS command other system
files. It is often a memory resident virus.
Compression – appended to executables
Companion virus - A specific type of virus where the
infected code is stored not in the host program, but in a
separate ‘companion’ files. For example, the virus might
rename the standard NOTEPAD.EXE file to NOTEPAD.EXD
and create a new NOTEPAD.EXE containing the virus code.
When the user subsequently runs the Notepad application,
the virus will run first and then pass control to the original
program, so the user doesn’t see anything suspicious. Takes
advantage of search order of an OS
Stealth virus – hides modifications to files or boot records
and itself
Multipart virus - infects both the boot sector and executable
files; becomes resident first in memory and then infects the
boot sector and finally the entire system
Self-garbling virus – attempts to hide by garbling its code;
as it spreads, it changes the way its code is encoded
Polymorphic virus – this is also a self-garbling virus where
the virus changes the "garble" pattern each time is spreads.
As a result, it is also difficult to detect.
Macro virus – usually written in Word Basic, Visual Basic or
VBScript and used with MS Office
Resident virus – Virus that loads when a program loads in
memory
Non-resident virus - attached to .exe
ANTI-Virus
Signature based cannot detect new malware
Heuristic behavioral can detect new malware
Programs
Compiler Translates higher level program into an
executable file
Interpreter reads higher level code, one line at the time to
produce machine instructions
Assembler converts machine-code into binary machine
instructions. Translate assembly language into machine
language.
Software Life Cycle
Requirements
Design
Programming
Testing
Conversion
Operations
Maintenance
More things to know
Black-box testing observes the system external behavior.
White-box testing is a detailed exam of a logical path,
checking the possible conditions.

Compiled code poses more risk than interpreted code
because malicious code can be embedded in the compiled
code and can be difficult to detect.

Regression testing is the verification that what is being
installed does not affect any portion of the application system
already installed. It generally requires the support of
automated process to repeat tests previously undertaken.

Code comparison is normally used to identify the parts of the
source code that have changed.

Integration testing is aimed at finding bugs in the relationship
and interfaces between pairs of components. It does not
normally test all functions.

Unit testing is the testing of a piece of code. It will only detect
errors in the piece of code being tested.
Control
Accuracy
Security
Consistency
Labels, traffic DBMS, data
Preventative
Data checks,
dictionary
padding,
validity
encryption
checks
Detective
Cyclic
IDS, audit
Comparison
Redundancy
trails
tools
Corrective
Checkpoint,
Emergency
Database
backups
response
controls


Domain 8 – Business Continuity & Disaster Recovery
DRIVERS
-
-
Business need to minimize loss.
o
Online service providers like Google and
EBay and NYSE need to be online
o
Retain value. Customer records = high value,
lost data reduces brand quality
Regulatory compliance
o
Utility companies (gas energy water)
o
Government (FISMA, NIST)
o
Finance (sox, FFIEC Basel II)
o
Healthcare (HIPAA)
Roles and responsibilities
BCP committee
Senior staff (ultimate responsibility, due care/diligence)
Various business units (identify and prioritize time
critical systems)
Information Systems
Security Administrator
People who will carry out the plan (execute)
Representatives from all departments
BIA
Goal: to create a document to be used to help understand what
impact a disruptive event would have on the business
-
-
Role of IT department
BCP
-
Plan for emergency response, backup operations and postdisaster recovery maintained by an activity as a part of its security
program that will ensure the availability of critical resources and
facilitate the continuity of operations in an emergency situation
NIST
3 Phases of actions
Notification/activation
Recovery
Reconstitution (back up and running)
-
Make sure that adequate backup restore processes are
available, including off-site media storage
Employ sufficient physical security mechanisms to
protect network and hardware components
Ensure that the organization uses sufficient logical
security measures for protecting sensitive data
Ensure that departments implements adequate system
administration, including up-to-date inventories of
hardware, software and media storage
-
-
BCP goals
BCP&DRP Goals
Business continuity- Ensuring the business can continue in an
emergency
Focus on business processes
Scope/plan initiation
Part of your security program
Need for management support
BIA – business impact analysis
BCP Development
At least once a year testing
Disaster
-
Recovery – Recover as quickly as possible
Heavy IT focus
Allows the execution of the BCP
Needs Planning
Needs Testing
Threats
-
Prevent interruption to normal business activity and
critical business process
Minimize the effects of a disaster
Quick recovery of all business
Should cover:
o
#1 PEOPLE (ALWAYS FIRST)
o
LAN/Wan
o
Telecom, data links
o
Workstations and workspace
o
Application software and DATA
o
Media and record storage
o
Staff duties
Business Continuity plans development
-
-
BCP elements
1.
2.
3.
Natural (Fires, explosions water, storm)
Man made (bombing, strikes, toxin spills)
4.
Scope and plan initiation - Consider amount of work
required, resources required, management practice
BIA – helps to understand impact of disruptive
processes
Business Continuity Plan development
a. Use BIA to develop BCP
b. Testing
Plan approval and implementation
Management approval
Create awareness
Update plan as needed
Gathering assessment material
o
Org charts to determine functional
relationships
o
Examine business success factors
Vulnerability assessment
o
Identify Critical IT resources out of critical
processes
o
Identify disruption impacts and Maximum
Tolerable Downtime (MTD)
o
Loss Quantitative (revenue, expenses for
repair) or qualitative (competitive edge, public
embarrassment). Presented as low, high,
medium.
o
Develop recovery procedures
Analyze the compiled information
o
Document the process
o
Identify inter-dependability
o
Determine acceptable interruption periods
Documentation and Recommendation
o
Presentation to management
Defining the continuity strategy
o
Computing: strategy to preserve the elements
of hardware/software/communication
lines/applications/data
o
Facilities: use of main buildings or any remote
facilities
o
People: operators, management, technical
support persons
o
Supplies and equipment: paper, forms HVAC
Documenting the continuity strategy
Domain 8 – Business Continuity & Disaster Recovery
Disaster Recovery Planning
Statement of actions that have to be taken before, during and
after a disruptive event that causes a significant loss of
information Goal: provide organized way for decision making,
reduce confusion and deal with the crisis. Planning and
development must occur before the disaster
BIA has already been done, now were going to protect!
Disaster Planning Process
-
Disaster Processing Continuity plan
Disaster Recovery plan maintenance
Service bureaus
Contract with a service bureau to fully provide alternate backup
processing services. Advantage: quick response and availability,
testing is possible. Disadvantage: expense and it is more of a
short time option.
Other data center backup alternatives
Rolling/mobile sites. Mobile homes or HVAC trucks.
Could be considered a cold site
In-house or external supply of hardware replacements.
Stock of hardware either onsite or with a vendor. May
be acceptable for warm site but not for hot site.
Prefabricated buildings. A very cold site.
Disaster Processing Continuity plan
Mutual aid agreements (aka reciprocal agreement)
Arrangement with another similar corporation to take over
processes. Advantage: cheap. Disadvantage: must be exact the
same, is there enough capability, only for short term and what if
disaster affects both corporations. Is not enforceable.
Subscription services
Third party, commercial services provide alternate backups and
processing facilities. Most common of implementations!
HOT SITE Fully configured computer facility. All
applications are installed, up-to-date mirror of the
production system. For extremely urgent critical
transaction processing. Advantage: 24/7 availability and
exclusive use are assured. Short and long term.
Disadvantage: extra administrative overhead, costly,
security controls needs to be installed at the remote
facility too. Exclusive to one company
WARM SITE Cross between hot and cold site. The
computer facility is available but the applications may
not be installed or need to be configured. External
connections and other data elements that take long time
to order are present. Workstations have to be delivered
and data has to be restored. Advantage: Less costly,
more choices of location, less administrative resources.
Disadvantage: it will take some time to start production
processing. Nonexclusive.
COLD SITE Least ready but most commonly used. Has
no hardware installed only power and HVAC.
Disadvantage: Very lengthy time of restoration, false
sense of security but better than nothing. Advantage:
Cost, ease of location choice. Nonexclusive
Multiple centers (aka dual sites)
Processing is spread over several computer centers. Can be
managed by same corporation (in-house) or with another
organization (reciprocal agreement). Advantage: costs, multiple
sites will share resources and support. Disadvantage: a major
disaster could affect both sites; multiple configurations have to be
administered.
Transaction Redundancy Implementations
Electronic vaulting transfer of backup data to an offsite storage
location via communication lines
Remote Journaling parallel processing of transactions to an
alternative site via communication lines
Database shadowing live processing of remote journaling and
creating duplicates the database sets to multiple servers
Disaster recovery plan test types
1.
2.
3.
4.
5.
Checklist test copies of the plan are distributed to
management for review
Structured Walk-Through test business unit
management meets to review the plan
Simulation test all support personnel meet in a
practice room
Parallel test Critical systems are run at an alternate site
Full-Interruption test Normal production shut down,
with real disaster recovery processes
Backup types
Full All files, archive bit and modify bit are cleared. Advantage:
only previous day needed for full restore, disadvantage: time
consuming
Incremental only modified files, archive bit cleared, Advantage:
least time and space, Disadvantage: first restore full then all
incremental backups, thus less reliable because it depends on
more components
Differential: only modified files, doesn’t clear archive bit.
Advantage: full and only last diff needed, Intermediate time
between full and diff.
Disaster recovery process
TEAMS
Recovery team mandated to implement recovery after the
declaration of the disaster
Salvage team goes back to the primary site to normal processing
environmental conditions. Clean, repair, Salvage. Can declare
when primary site is available again
Normal Operations Resume plan has all procedures on how the
company will return processing from the alternate site
Other recovery issues
Interfacing with other groups: everyone outside the corporation
Employee relations: responsibility towards employees and their
families
Fraud and Crime: like vandalism, looting and people grabbing the
opportunity
Financial disbursement
Media relations
Things to know
The disaster is not over until all operations have been returned to
their normal location and function
It will be officially over when the data has been verified at the
primary site, as accurate
RTO: recovery time objectives. Refers to business processes not
hardware.
RTO 5 minutes or hours Hot site; RTO 1-2 days  warm site
RTO 3-5 days  mobile site; RTO 1-2 weeks cold site
Backup storage media
Tape: sequential, slow read, fast write 200GB an hour, historically
cheaper than disk (now changing), robotic libraries
Disk fast read/write, less robust than tape
Optical drive: CD/DVD. Inexpensive
Solid state: USB drive, security issues
MTTF (mean time to failure)
MTTR (mean time to repair)
MTBF Mean time between failures (Useful Life) = MTTF + MTTR
RPO -Recovery Point Objective: Point in time that application data
must be recovered to resume business functions
MTD -Maximum Tolerable Downtime: Maximum delay a business
can be down and still remain viable
MTD minutes to hours: critical
MTD 24 hours: urgent
MTD 72 hours: important
MTD 7 days: normal
MTD 30 days non-essential
Domain 9 – Law, Investigation and Ethics
Terms
Ethics
Intellectual property laws
Wire Tapping eavesdropping on communication -only legal with
prior consent or warrant
Dumpster Driving act of going through someone's trash to find
useful or confidential info -it is legal but unethical in nature
Phishing act of sending spoofed messages that pretend to
originate from a source the user trusts (like a bank)
Social Engineering act of tricking someone into giving sensitive
or confidential info that may be used against the company
Script kiddie someone with moderate hacking skills, gets code
from the Internet.
Data Diddling act of modifying information, programs, or
documents to commit fraud, tampers with INPUT data
Privacy Laws data collected must be collected fairly and lawfully
and used only for the purpose it was collected.
Computer Crime Laws -3 types of harm
unauthorized intrusion,
unauthorized alteration or destruction
malicious code
Admissible evidence relevant, sufficient, reliable
Red boxing: pay phones cracking
Black Boxing manipulates toll-free line voltage to phone for free
Blue Boxing tone simulation that mimics telephone co. system
and allows long distance call authorization
Phreakers hackers who commit crimes against phone companies
Salami removal of a small amount of money otherwise known as
skimming
Hearsay second-hand data not admissible in court
Federal Sentencing provides judges and courts procedures on
Guidelines the prevention, detection and reporting of crimes that
should occur by a company official and made company executives
responsible for the company's actions
Due Care
Which means when a company did all that it could have
reasonably done to try and prevent security breach / compromise /
disaster, let’s call it damage and took the necessary steps required
as countermeasures / controls, let’s call it safeguards. The benefit
of "due care" can be seen as the difference between the damage
with or without out "due care" safeguards in place. AKA doing
something about the threats
Due Diligence
means that the company properly investigated all of its possibly
weaknesses and vulnerabilities AKA understanding the threats
Enticement is the legal action of luring an intruder, like in a honeypot
Entrapment is the illegal act of inducing a crime, the individual
had no intent of committing the crime at first
Just because something is legal doesn’t make it right.
Within the ISC context: Protecting information through CIA
Patent grants ownership of an invention and provides enforcement
for owner to exclude others from practicing the invention. After 20
years the idea is open source
Copyright protects the expression of ideas but not necessarily the
idea itself
Trade Secret something that is propriety to a company and
important for its survival and profitability (like formula of Coke or
Pepsi)
Trademarks words, names, product shape, symbol, color or a
combination used to identify products and distinguish them from
competitor products (McDonald’s M)
Five rules of evidence:

Be authentic

Be accurate

Be complete

Be convincing

Be admissible
Code of Ethics Canons:
–Protect society, the commonwealth, and the infrastructure.
–Act honorably, honestly, justly, responsibly, and legally.
–Provide diligent and competent service to principals.
–Advance and protect the profession.
Internet Advisory Board (IAB)
Ethics and Internet (RFC 1087):
Access to and use of Internet is a privilege and should be
treated as such
It is defined as unacceptable and unethical if you for example
gain unauthorized access to resources on the internet, destroy
integrity waste resources or compromise privacy.
Corporate Officer Liability
•Executives are now held liable if the organization they
represent is not compliant with the law.
•Negligence occurs if there is a failure to implement
recommended precautions, if there is no contingency/disaster
recovery plan, failure to conduct appropriate background
checks, failure to institute appropriate information security
measures, failure to follow policy or local laws and regulations.
Law
Common law: USA, UK Australia Canada (judges)
Civil law: Europe, south America
Islamitic and other Religious laws: middle east Africa
Indonesia
USA
3 branches for laws:
Legislative: writing laws (statutory laws).
Executive: enforces laws (administrative laws)
Juridical: Interprets laws (makes common laws out of court
decisions)
3 categories
Criminal law – individuals that violate government laws.
Punishment mostly imprisonment
Civil law – wrongs against individual or organization that result
in a damage or loss. Punishment can include financial
penalties. AKA tort law (I’ll Sue You!) Jury decides upon
liability
Administrative/Regulatory law – how the industries,
organizations and officers have to act. Wrongs can be
penalized with imprisonment or financial penalties
Incident Response
Events: anything that happens. Can be documented verified and
analyzed
Incident: event or series of events that adversely impact the ability of
an organization to do business
Framework: Response Capability (policy, procedures, a team),
Incident response and handling (Triage, investigation,
containment, and analysis & tracking), Recovery (Recovery /
Repair), Debriefing / Feedback (External Communications, Metrics)
Regulations
SOX 2002 after ENRON and World Online debacle
Independent review by external accountants.
Section 302: CEO’s CFO’s can be sent to jail when information they
sign is incorrect.
Section 404 is the about internal controls assessment: describing
logical controls over accounting files; good auditing and information
security.
European laws:
Need for information security to protect the individual.
Privacy is the keyword here! Only use information of individuals for
with it was gathered
(remember ITSEC, the European version of TCSEC that came from
the USA/orange book, come together in Common criteria, but there
still is some overlap)

strong in anti spam and legitimate marketing

Directs public directories to be subjected to tight controls

Takes an OPT-IN approach to unsolicited commercial
electronic communications

User may refuse cookies to be stored and user must be
provided with information

Member states in the EU can make own laws of e.g.
retention of data
Domain 9 – Law, Investigation and Ethics
Evidence
Witnesses
Investigation
Sufficient –persuasive enough to convince one of the validity of
the findings
Reliable –consistent with fact
Relevant –relationship to the findings must be reasonable and
sensible
Permissible – lawful obtaining of evidence
Preserved and identifiable – collection, reconstruction
EVIDENCE LIFECYCLE
1.
Discovery
2.
Protection
3.
Recording
4.
Collection and identification
5.
Analysis
6.
Storage, preservation, transportation
7.
Present in court
8.
Return to owner
Opinion Rule
–Requires witnesses to testify only about the facts of the case,
cannot be used as evidence in the case.
Expert Witnesses
–Used to educate the jury, can be used as evidence.
MOM means, opportunity and motive
Determine suspects
Victimology –why certain people are victims of crime and how
lifestyle affects the chances that a certain person will fall victim to a
crime Investigation
Target Risk Assessment –why was target chosen –history of target
•Crime scene characteristics
•Attacker skill level
•Intent
When investigating a hard drive, don’t use message digest because
it will change the timestamps of the files when the file-system is not
set to Read-Only
Slack space on a disk should be inspected for hidden data and
should be included in a disk image
Types of evidence
Best Evidence:
–Primary Evidence–is used at the trial because it is the most
reliable.
–Original documents–are used to document things such as
contracts –NOTE: no copies!
–Note: Oral is not best evidence though it may provide
interpretation of documents, etc.
Secondary Evidence
–Not as strong as best evidence.
–A copy, Secondary Evidence, is not permitted if the original, Best
Evidence, is available
–Copies of documents.
–Oral evidence like Witness testimony
Direct Evidence:
–Can prove fact by itself and does not need any type of backup
information.
–Testimony from a witness –one of their 5 senses:
•Oral Evidence is a type of Secondary Evidence so the case can’t
simply stand on it alone
•But it is Direct Evidence and does not need other evidence to
substantiate it
Conclusive evidence
–Irrefutable and cannot be contradicted
–Requires no other corroboration
Circumstantial evidence
–Used to help assume another fact
–Cannot stand on its own to directly prove a fact
Corroborative Evidence:
–Supports or substantiates other evidence presented in a case
Hearsay Evidence something a witness hear another one say.
Also business records are hearsay and all that’s printed or
displayed. One exception to business records: audit trails and
business records are not considered hearsay when the documents
are created in the normal course of business.
Admissibility of evidence
For evidence to be admissible it must be:
Relevant
•Proof of crime, documentation of events, proof of acts and
methods used, motive proof, identification of acts
Legally permissible obtained in a lawful manner
•Avoid: unlawful search and seizure, secret recording, privacy
violations, forced confessions, unlawful obtaining of evidence
Reliable evidence has not been tampered with or modified
Identification labeling, recording serial number etc.
Evidence must be preserved and identifiable
•Collection, documentation, classification, comparison,
reconstruction
•Witnesses that evidence is trustworthy, description of
procedures, normal business methods collections, error
precaution and correction
Laws
1974 US Privacy Act: Protection of PII on federal databases
1980 Organization for Economic Cooperation and
Development (OECD): Provides for data collection,
specifications, safeguards
1986 (amended in 1996) US Computer Fraud and Abuse
Act: Trafficking in computer passwords or information that
causes a loss of $1,000 or more or could impair medical
treatment.
1986 Electronic Communications Privacy Act: Prohibits
eavesdropping or interception w/o distinguishing private/public
1987 US Computer Security Act: Security training, develop a
security plan, and identify sensitive systems on govt agencies.
1991 US Federal Sentencing Guidelines: Responsibility on
senior management with fines up to $290 million. Invoke
prudent man rule. Address both individuals and organizations
1996 US Economic and Protection of Propriety
Information Act: industrial and corporate espionage
1996 Health Insurance and Portability Accountability Act
(HIPPA)
1996 US National Information Infrastructure Protection
Act: Encourage other countries to adopt similar framework.
Interstate commerce clause: Federal government has power
to regulate all trade between states.
Interviewing and Interrogation
Interviewing –To ultimately obtain a confession, discover
information
Interrogation–Evidence retrieval method
The Process
–Prepare questions and topics, put witness at ease, summarize
information –interview/interrogation plan
–Have one person as lead and 1-2 others involved as well –never
interrogate or interview alone
Things to know
Hackers and crackers want to verify their skills as intruders
Notebook: most preferred in the legal investigation is a bound
notebook, pages are attached to a binding.
Exigent circumstances allows officials to seize evidence before its
destroyed (police team fall in)
Data haven is a country or location that has no laws or poorly
enforced laws
Chain of custody = collection, analysis and preservation of data
Forensics uses bit-level copy of the disk
Residual risk = where cost of applying extra countermeasures is
more than the estimated loss resulting from a threat or vulnerability
(C > L). Legally the remaining residual risk is not counted when
deciding whether a company is liable.
FAIR INFORMATION PRACTICES

Openness

Collection Limitation

Purpose Specification

Use Limitation

Data Quality

Individual Participation

Security Safeguards

Accountability
THREATS
Natural environment threats (earthquakes floods, tornadoes)
Supply system threats (power communications water gas)
Manmade threats (vandalism, fraud, theft)
Politically motivated threats (terroristic attacks, riots bombings)
Life safety takes precedence!!
Layered defense model: all physical controls should be work
together in a tiered architecture (stacked layers)
Vulnerability=weakness threat = someone will identify the
weakness and use it against you and becomes the threat agent
Domain 10 - Physical Security
Risk analysis-->Acceptable risk level -->baseline>implement
countermeasures
Major sources:
Temperature
Gases
Liquids
Organism: viruses, bacteria
Projectiles: cars, trucks, bullets
Movement: Collapse, earthquakes
Energy: radio, radiation
CONTROLS
Physical (Fences Trees Locks)
Administrative (badges clothing procedures)
Technical (Alarms Humidity AC heat cameras)
TYPES OF CONTROL
Preventive (guards dogs firewall)
Detective (CCTV, motion detectors audit logs)
Corrective (IDS Antivirus)
Deterrents (fences, alarms personnel)
Recover (backup)
Compensating (monitoring supervising)
Electrical power
Interference
Clean=no interference
Line noise: can be EMI or RFI
Transient: short duration of noise
Counter: voltage regulators, grounding/shielding and line
conditioners
EMI
COMMON mode noise: difference between hot and ground
Traverse mode noise: difference between hot and neutral
HINT: common--grounds
Excesses
SPIKE: short high voltage
SURGE: long high voltage
Counter: surge protector
Losses
FAULT: short outage
BLACKOUT: long outage
Counter: Backup power
Long term: Backup Power generator
Short term: UPS
-Online uses ac line voltage to charge batteries, power always
though UPS
-Standby UPS, inactive till power down
Degradation
SAG/DIP: short low voltage
BROWNOUT: long low voltage
Counter: constant voltage transformers
Other
Inrush Surge: surge of current required to power on devices
Common-mode noise: radiation from hot and ground wires
Traverse-mode noise: radiation from hot and neutral wires.
Static charge
40 sensitive circuits
1000 scramble monitor display
1500 disk drive data loss
2000 system shutdown
4000 Printer Jam
17000 Permanent chip damage
Use antistatic spray and flooring, ground rooms properly
Humidity
<40% static electricity up to 20.000 volts
NORMAL 40-60% up to 4000 volts
>60% corrosion
Fire
Prevention
Training construction, supplies, reach ability
Detection
Manual: pull boxes
Automatic dial- up: Fire department, aka Auxiliary station alarm
Detectors:
Smoke activated,
Heat activated,
Flame activated(infrared)
Classes
A Common WATER, SODA ACID
B Liquids----GAS/CO2, SODA ACID
C Electrical-----GAS/CO2
D Metals----DRY POWDER
WATER suppress temperature
SODA ACID reduces fuel supply
CO2 reduces oxygen
HALON chemical reaction
Fire distinguishers should be 50 feet from equipment and toward
the door
Heat
Computer hardware 175F (80c)
Magnetic storage 100F (37c)
Paper 350F (176c)
Sprinklers
Wet pipe
always contains water, fuse nozzle melts at 165F
Dry pipe
water in tank until clapper valve releases it
Deluge
Douches, large amounts of water/foam
Pre-action (MOST RECOMMENDED)
water in tanks, first water in pipes when air is lost when heat is
detected, then thermal link in nozzle melts to release water
HALON
1211 = portable
1301 = flooding
FM-200 most common replacement (others: CEA, NAF, FE-13
Argon INERGEN Low Pressure Water)
RESISTANCE
Walls: 1 hour fire rating and adjacent room with paper 2 hours
Locks
Warded lock hanging lock with a key
Tumbler lock cylinder slot
Combination lock 3 digits with wheels
Cipher Lock Electrical
Device lock bolt down hardware
Preset ordinary door lock
Programmable combination or electrical lock
Raking = circumvent a pin tumbler lock
Domain 10 - Physical Security
Lightning
Glare protection against blinding by lights
Continuous lightning evenly distributed lightning
Controlled lightning no bleeding over no blinding
Standby Lightning timers
Responsive areas illumination IDS detects activities and turns on
lightning
NIST: for critical areas the area should be illuminated 8 feet in
height with 2-foot candle power
Fences
Small mesh and high gauge is most secure
3-4 feet deters casual trespasser
6-7 feet to hard to climb easily
8 feet + wires deters intruders,
no one STOPS a determined intruder
Location
CPTED Crime Prevention Through Environmental design
Natural Access control: guidance of people by doors
fences bollards lightning. Security zones defined
Natural surveillance: cameras and guards
Territorial Reinforcements: walls fences flags
Target Hardening: focus on locks, cameras guards
rd
Facility site: CORE OF BUILDING (thus with 6 stores, on 3 floor)
FAIL SAFE: doors UNLOCK
FAIL SECURE: doors LOCK
CCTV
Multiplexer allows multiple camera screens shown over one cable
on a monitor
Via coax cables (hence closed)
Attacks: replayed (video images)
Fixed mounting versus PTZ Pan Tilt Zoom
accunicator system (detects movements on screen and alerts
guards)
Recording (for later review) = detective control
Intrusion detection
PHYSICAL PARAMETER DETECTION
Electromechanical: detect a break or change in a circuit
magnets pulled lose, wires door, pressure pads
Photoelectric: light beams interrupted (as in an store entrance)
Passive infrared: detects changes in temperature
acoustical detection: microphones, vibrations sensors
MOTION
wave pattern motion detectors: detects motions
proximity or capacitance detector: magnetic field detects presence
around an object
ALARMS
Local alarms audible alarm for at least 4000 feet far
Central stations less than 10mins travel time for e.g. an private
security firm
Proprietary systems owned and operated by the customer.
System provides many of the features in-house
Auxiliary Station systems on alarm ring out to local fire or police
Line supervision check if no tampering is done with the alarm
wires
Power supplies alarm systems needs separate circuitry and
backup power
Data destruction and reuse
Object reuse: use after initial use
Data remanence: remaining data after erasure
Format magnetic media 7 times (orange book)
Clearing: overwriting media to be reused
Purging: degaussing or overwriting to be removed
Destruction: complete destroy preferably by burning
Audit trails
Date and time stamps
Successful or not attempt
Where the access was granted
Who attempted access
Who modified access privileges at supervisor level
Security access cards
Photo id card: dumb cards
Digital-coded cards:

Swipe cards

Smartcards
Wireless proximity cards

User activated

System sensing
o
Passive device, no battery, uses power of the
field
o
Field Powered device: active electronics,
transmitter but gets power from the
surrounding field from the reader
o
Transponders: both card and receiver holds
power, transmitter and electronics
Other things to know
Piggybacking: looking over someone’s shoulder to see how
someone gets access.
Data center should have:

Walls from floor to ceiling

Floor: Concrete slab: 150 pounds square foot

No windows in a datacenter

Air-conditioning should have own Emergency Power Off
(EPO)
Electronic Access Control (EAC): proximity readers,
programmable locks or biometric systems
Order of



actions when fire is detected:
Evacuate the facility
Shut down computer systems and power if possible
Inform facility management contract fire department